{"id":8103,"date":"2026-08-29T15:00:17","date_gmt":"2026-08-29T15:00:17","guid":{"rendered":"https:\/\/pishi.ai\/?post_type=docs&#038;p=8103"},"modified":"2026-10-08T10:09:53","modified_gmt":"2026-10-08T10:09:53","password":"","slug":"object-detection-detect-objects-in-scratch","status":"publish","type":"docs","link":"https:\/\/pishi.ai\/ar\/help\/object-detection-detect-objects-in-scratch\/","title":{"rendered":"\u0643\u0634\u0641 \u0627\u0644\u0643\u0627\u0626\u0646\u0627\u062a - \u0627\u0643\u062a\u0634\u0627\u0641 \u0627\u0644\u0643\u0627\u0626\u0646\u0627\u062a \u0641\u064a \u0633\u0643\u0631\u0627\u062a\u0634"},"content":{"rendered":"<h2>\ud83c\udfaf Object Detection &#8211; Detect Objects in Scratch<\/h2>\n<p>The Object Detection extension brings real AI-powered object recognition into Scratch.<br \/>\nIt lets your project identify multiple objects from the live camera or Scratch stage, count them, read their position and size, and react when objects appear.<br \/>\nBecause the models are already trained, students can start experimenting with computer vision immediately &#8211; no model training or external setup required.<br \/>\nSimple enough for beginners, but powerful enough for classroom AI, STEM, and interactive game projects. \ud83e\udd16<\/p>\n<p><img fetchpriority=\"high\" decoding=\"async\" class=\"help-intro-image\" src=\"https:\/\/pishi.ai\/wp-content\/uploads\/2026\/09\/pishi-ai-object-detection-scratch.jpg\" alt=\"Pishi points to a screen detecting and labeling an apple, bottle, cup, backpack, ball, and toy.\" width=\"400\" height=\"400\" \/><\/p>\n<hr \/>\n<h2>\ud83c\udf1f Overview<\/h2>\n<ul>\n<li><strong>Detect Multiple Objects:<\/strong> Recognize several supported objects in the same image or camera frame.<\/li>\n<li><strong>Count What the AI Sees:<\/strong> Read the total number of detections or count only a specific object.<\/li>\n<li><strong>Track Position and Size:<\/strong> Get the X\/Y center, width, and height of detected objects in Scratch stage units.<\/li>\n<li><strong>Filter What Matters:<\/strong> Detect everything the model recognizes, or create your own list of object names to focus on.<\/li>\n<li><strong>Save and Sort Detections:<\/strong> Save a snapshot of selected detections and order them by confidence, position, or size.<\/li>\n<li><strong>Trigger Events:<\/strong> Use hat blocks to react when an object appears or when the number of detections changes.<\/li>\n<li><strong>Choose the Right Model:<\/strong> Switch between Fast, Balanced, and Wide detection options depending on your device, image detail, and the variety of objects you need.<\/li>\n<li><strong>Choose the Input:<\/strong> Detect objects from the live camera or directly from the Scratch stage image.<\/li>\n<\/ul>\n<hr \/>\n<h2>\u2728 Key Features<\/h2>\n<ul>\n<li>Pre-trained, prompt-free object detection &#8211; no training step required.<\/li>\n<li>Detect many objects in one frame and work with each detection individually.<\/li>\n<li>Read object names, counts, confidence, position, width, and height.<\/li>\n<li>Choose between current live detections and a saved snapshot when building more advanced projects.<\/li>\n<li>Create a custom detection list to ignore objects that are not important to your project.<\/li>\n<li>Save and sort detections from left to right, top to bottom, largest to smallest, or highest to lowest confidence &#8211; or reverse any of these orders.<\/li>\n<li>Hat blocks make it easy to build object-controlled games and interactive scenes.<\/li>\n<li>Adjustable detection intervals help balance speed and device performance.<\/li>\n<li>Works fully in-browser &#8211; <strong>safe and private<\/strong>.<\/li>\n<\/ul>\n<hr \/>\n<h2>\ud83d\ude80 How to Use<\/h2>\n<ol>\n<li>Go to: <a href=\"https:\/\/pishi.ai\/play\" target=\"_blank\" rel=\"noopener\">pishi.ai\/play<\/a><\/li>\n<li>Open the Extensions section.<\/li>\n<li>Select the Object Detection extension.<\/li>\n<li>Allow camera access when prompted and check that the preview appears.<\/li>\n<li>If you prefer, use the Scratch stage as the input to detect objects in the stage image, including any camera video displayed there. For the best real-time performance, use Camera input, since processing the Stage may be slower.<\/li>\n<li>The extension starts with Fast detection model. You can switch to Fast, Balanced, or Wide at any time to change the balance between speed, detail, and object coverage.<\/li>\n<li>Try the &#8220;list of [current] detections&#8221; reporter block to see the object names the AI recognizes.<\/li>\n<li>Then use counting, position, saved-detection, filtering, and event blocks to build your project.<\/li>\n<\/ol>\n<p><strong>Tips<\/strong><\/p>\n<ul>\n<li>Use clear lighting and keep the object large enough to see easily in the frame.<\/li>\n<li>Start with one supported object before building a project that depends on several objects.<\/li>\n<li>Use the exact object label returned by the &#8220;list of [SOURCE] detections&#8221; reporter block when typing an object name into another block.<\/li>\n<li>The object vocabulary and detection labels are in English, so use English labels when entering object names.<\/li>\n<li>current means the latest live detection results. saved means a snapshot created with the save-and-order block.<\/li>\n<li>If the same kind of object appears more than once, use #1, #2, #3, and so on to work with individual detections.<\/li>\n<li>If detection feels slow on a device, try the Fast model or increase the classification interval.<\/li>\n<\/ul>\n<hr \/>\n<h2>\ud83e\uddf1 Blocks and Functions<\/h2>\n<p>&nbsp;<\/p>\n<h3>\ud83d\udcca Detection Lists, Sources, and Counts<\/h3>\n<p><strong>Understanding SOURCE:<\/strong><\/p>\n<ul>\n<li><strong>current<\/strong> &#8211; the latest live detections from the camera or stage.<\/li>\n<li><strong>saved<\/strong> &#8211; a snapshot stored by this block:<br \/>\n\t\t<span id=\"block-save-and-order-detections\" class=\"scratchblocks\">save [OBJECT|text] from [SOURCE] detections, order by [ORDER_TYPE]<\/span>\n\t<\/li>\n<\/ul>\n<p>Using a saved snapshot is helpful when you want object order and positions to stay fixed while your project works with them.<\/p>\n<p>&nbsp;<\/p>\n<p><strong>Shared behavior:<\/strong> Reporter blocks that return a name, list, confidence, position, or size return an empty value when there is no matching detection, no saved snapshot is available, or the requested item number does not exist. Count reporters return 0, and Boolean blocks return false.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-list-of-detections\" class=\"scratchblocks\">list of [SOURCE] detections<\/div>\n<p>Reports the object labels in the selected detection source as comma-separated text.<br \/>\nChoose current to see what the AI recognizes right now, or saved to read the stored snapshot.<br \/>\nThis is one of the best blocks for learning the exact object names the selected model uses in its detection results.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-object-count\" class=\"scratchblocks\">number of [SOURCE] detections<\/div>\n<p>Reports the total number of objects in the selected source.<br \/>\nFor example, it can count all live current detections or tell you how many objects are stored in saved detections.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-all-items-argument\" class=\"scratchblocks\">all<\/div>\n<p>Reports the special value all, which represents every detected object regardless of its label.<br \/>\nWhen you want a block to work with every detected object instead of one specific label, place it in an input of a block that asks for an object name (<span class=\"scratchblocks\" data-scratch-input-type=\"text\">OBJECT<\/span>).<br \/>\nYou can also type all in English directly into the object-name input instead of using this reporter block.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-save-and-order-detections\" class=\"scratchblocks\">save [OBJECT|text] from [SOURCE] detections, order by [ORDER_TYPE]<\/div>\n<p>Takes matching detections from the selected source, sorts them, and stores the result as the new saved snapshot.<br \/>\nThis is especially useful when you need a predictable order such as leftmost first, largest first, or highest confidence first.<\/p>\n<p>By default, the current detection list is already ordered from highest to lowest confidence.<\/p>\n<p><span class=\"scratchblocks\" data-scratch-input-type=\"text\">OBJECT<\/span>: enter an object label such as person, comma-separated text such as person, car, or use all.<\/p>\n<ul>\n<li><strong>all<\/strong>: keeps all detections and saves them in the selected order. If you choose confidence high to low, the current detections are already in that order, so the block simply saves them as a snapshot.<\/li>\n<li><strong>An object name, such as person<\/strong>: filters the detections to that object and then sorts the matching results.<\/li>\n<li><strong>Multiple object names, such as person, car<\/strong>: combines detections matching any listed name and sorts them together.<\/li>\n<\/ul>\n<p><span class=\"scratchblocks\" data-scratch-input-type=\"round\">SOURCE<\/span>: choose current or saved.<\/p>\n<ul>\n<li><strong>current<\/strong>: uses the latest live detection results.<\/li>\n<li><strong>saved<\/strong>: uses the existing saved snapshot, allowing you to reorder it or filter it further.<\/li>\n<\/ul>\n<p><span class=\"scratchblocks\" data-scratch-input-type=\"round\">ORDER_TYPE<\/span>: order type can be:<\/p>\n<ul>\n<li><strong>confidence high to low<\/strong> or <strong>confidence low to high<\/strong><\/li>\n<li><strong>left to right<\/strong> or <strong>right to left<\/strong><\/li>\n<li><strong>top to bottom<\/strong> or <strong>bottom to top<\/strong><\/li>\n<li><strong>large to small<\/strong> or <strong>small to large<\/strong><\/li>\n<\/ul>\n<p>After saving, choose saved in the SOURCE menu of other blocks to read the snapshot.<br \/>\nSaved detections stay unchanged until you save again, and switching the detection model clears the saved snapshot.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-nameof-detection-by-index\" class=\"scratchblocks\">name of #[INDEX|text] [SOURCE] detection<\/div>\n<p>Reports the label of one detection by its position in the selected source.<br \/>\n<span class=\"scratchblocks\" data-scratch-input-type=\"text\">INDEX<\/span>: use 1 for the first item, 2 for the second, and so on.<br \/>\nIf the item does not exist, the block returns empty.<br \/>\nFor saved detections, item 1, item 2, and so on follow the order that was selected when the detections were saved.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-number-of-detections-by-name\" class=\"scratchblocks\">number of [OBJECT|text] in [SOURCE] detections<\/div>\n<p>Reports how many detections in the selected source match an object name.<br \/>\n<span class=\"scratchblocks\" data-scratch-input-type=\"text\">OBJECT<\/span>: enter an exact object label, comma-separated text, or use all to count every object in the source.<br \/>\nWhen multiple names are provided, the block returns their combined detection count. For example, person, car counts every detected person and car.<br \/>\nUse this for counting games, classroom experiments, or logic such as \u201ccontinue when two people are visible.\u201d<\/p>\n<hr \/>\n<h3>\ud83d\udccd Position, Size, and Detection Confidence<\/h3>\n<div id=\"block-get-confidence-of\" class=\"scratchblocks\">confidence of [OBJECT|text] #[INDEX|text] in [SOURCE] detections<\/div>\n<p>Reports the confidence of a selected detection on a scale from 0 to 1.<br \/>\nFor example, 0.87 means the model gave that detection a confidence score of about 87%.<\/p>\n<p><span class=\"scratchblocks\" data-scratch-input-type=\"text\">OBJECT<\/span>: enter an exact object label, or use all to select items from the whole source.<\/p>\n<p><span class=\"scratchblocks\" data-scratch-input-type=\"text\">INDEX<\/span>: selects the first, second, third, and so on matching item.<\/p>\n<p><span class=\"scratchblocks\" data-scratch-input-type=\"round\">SOURCE<\/span>: choose current or saved.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-x-of\" class=\"scratchblocks\">x of [OBJECT|text] #[INDEX|text] in [SOURCE] detections<\/div>\n<p>Reports the X position of the center of the selected detection box, mapped to Scratch stage coordinates.<\/p>\n<p>Use it to move a sprite left or right with a real-world object.<\/p>\n<p><strong>Note:<\/strong> A detection box is the rectangular area around a detected object that shows where the object is located in the image.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-y-of\" class=\"scratchblocks\">y of [OBJECT|text] #[INDEX|text] in [SOURCE] detections<\/div>\n<p>Reports the Y position of the center of the selected detection box, mapped to Scratch stage coordinates.<br \/>\nUse it for vertical tracking, aiming, or object-controlled games.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-width-of\" class=\"scratchblocks\">width of [OBJECT|text] #[INDEX|text] in [SOURCE] detections<\/div>\n<p>Reports the width of the selected detection box in Scratch stage units.<br \/>\nA larger value means the detected object occupies more horizontal space in the image.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-height-of\" class=\"scratchblocks\">height of [OBJECT|text] #[INDEX|text] in [SOURCE] detections<\/div>\n<p>Reports the height of the selected detection box in Scratch stage units.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-is-anything-detected\" class=\"scratchblocks\">anything in [SOURCE] detections?<\/div>\n<p>Returns true when the selected source contains at least one detection and false when it is empty.<br \/>\nWith current, this checks the live scene. With saved, it checks the stored snapshot.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-is-object-detected\" class=\"scratchblocks\">[OBJECT|text] #[INDEX|text] detected in [SOURCE] detections?<\/div>\n<p>Checks whether a particular object and item number exists in the selected source.<br \/>\nFor example, person #2 is true only when a second matching person detection exists.<br \/>\nYou can also use all as the object value to test whether a particular item number exists in the whole source.<\/p>\n<hr \/>\n<h3>\ud83c\udfaf Detection Modes and Custom Lists<\/h3>\n<div id=\"block-set-look-for-mode\" class=\"scratchblocks\">look for [MODE]<\/div>\n<p>Controls which detected objects are included in the results used by the extension&#8217;s other blocks and detection events.<\/p>\n<p><span class=\"scratchblocks\" data-scratch-input-type=\"round\">MODE<\/span>:<\/p>\n<ul>\n<li><strong>all detections<\/strong> &#8211; keep every object recognized by the selected model.<\/li>\n<li><strong>my detection list<\/strong> &#8211; keep only object names that you have added to your custom detection list.<\/li>\n<\/ul>\n<p><strong>Important:<\/strong> The my detection list is empty when the extension starts. You can switch to this mode at any time, but no detections will be reported until the list contains at least one object name.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-my-detection-list\" class=\"scratchblocks\">my detection list<\/div>\n<p>Reports your custom object list as comma-separated text.<br \/>\nThe list can also store a history of newly detected object names without duplicates. For example, when the detection count changes, add the list of current detections to my detection list. Adding a name that is already in the list, even with different capitalization, does not add it again.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-my-detection-list-item\" class=\"scratchblocks\">my detection list item #[INDEX|text]<\/div>\n<p>Reports the object name at the selected position in your custom detection list.<br \/>\n<span class=\"scratchblocks\" data-scratch-input-type=\"text\">INDEX<\/span>: use 1 for the first item, 2 for the second, and so on.<br \/>\nIf the item does not exist, the block returns empty.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-get-my-detection-list-item-count\" class=\"scratchblocks\">number of items in my detection list<\/div>\n<p>Reports how many unique object names are in your custom detection list.<br \/>\nThis counts configured names, not objects currently visible to the camera.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-add-to-my-detection-list\" class=\"scratchblocks\">add [OBJECT|text] to my detection list<\/div>\n<p>Adds an object label to the custom list.<br \/>\nUse the exact label returned by the model.<br \/>\nComma-separated text adds each label separately.<br \/>\nAdding the same name again does not create a duplicate entry.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-remove-from-my-detection-list\" class=\"scratchblocks\">remove [OBJECT|text] from my detection list<\/div>\n<p>Removes the matching object name from your custom list.<br \/>\nComma-separated text removes every matching label.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-clear-my-detection-list\" class=\"scratchblocks\">clear my detection list<\/div>\n<p>Removes every name from the custom detection list.<\/p>\n<hr \/>\n<h3>\u26a1 Hat Blocks and Reactions<\/h3>\n<div id=\"block-when-any-object-detected\" class=\"scratchblocks\">when any object detected<\/div>\n<p>Runs the attached script when the stable number of any detected object increases.<br \/>\nIt can trigger when the first object appears, when a new object label appears, or when an additional instance of an already visible object is detected.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-when-object-detected\" class=\"scratchblocks\">when [OBJECT|text] detected<\/div>\n<p>Runs when the stable count of the named object increases.<br \/>\nFor example, it can trigger when the first person appears and again if another person is newly detected.<br \/>\nEnter comma-separated text to run when the stable count of any listed object increases.<br \/>\nUsing all here gives the same \u201cwhen any object detected\u201d behavior as the previous hat block.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-when-detections-count-changed\" class=\"scratchblocks\">when detections count changed<\/div>\n<p>Runs when the total stable number of currently detected objects changes, either up or down.<br \/>\nThis is useful for counting games and scenes that react when objects enter or leave the camera view.<\/p>\n<p>&nbsp;<\/p>\n<p><strong>Note:<\/strong> detection hats wait for the same count to be seen in 2 consecutive frames before treating it as stable. This helps reduce one-frame flicker and accidental triggers.<\/p>\n<hr \/>\n<h3>\ud83e\udde0 Model and Confidence Controls<\/h3>\n<div id=\"block-set-detection-model\" class=\"scratchblocks\">find using [MODEL_NAME] model<\/div>\n<p>Switches the active object detection model.<br \/>\nThe extension starts with the Fast model. Students can switch between models depending on whether their project needs faster detection, more image detail, or support for a wider range of object types.<\/p>\n<div class=\"pishi-table\">\n<table>\n<thead>\n<tr>\n<th>Model Menu<\/th>\n<th>Input Size<\/th>\n<th>Classes<\/th>\n<th>Best For<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><strong>fast size: 320 classes: 365<\/strong><\/td>\n<td>320\u00d7320<\/td>\n<td>365<\/td>\n<td>Lower processing load and smoother performance on less powerful devices<\/td>\n<\/tr>\n<tr>\n<td><strong>balanced size: 640 classes: 365<\/strong><\/td>\n<td>640\u00d7640<\/td>\n<td>365<\/td>\n<td>More image detail while using the same 365-class object vocabulary<\/td>\n<\/tr>\n<tr>\n<td><strong>wide size: 640 classes: 4585<\/strong><\/td>\n<td>640\u00d7640<\/td>\n<td>4,585<\/td>\n<td>A much wider object vocabulary when the 365-class options do not include what you need<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p class=\"pishi-table-note\">\u2194 Swipe left or right to view full table on mobile<\/p>\n<p>&nbsp;<\/p>\n<p><strong>Tip:<\/strong> Start with the default Fast model for responsive detection. Choose balanced for more image detail, or wide when you need many more object categories.<\/p>\n<p><strong>Important:<\/strong> switching models clears saved detections. The current minimum confidence setting is kept when switching models.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-set-minimum-confidence\" class=\"scratchblocks\">set minimum confidence [CONFIDENCE|text]<\/div>\n<p>Sets the minimum confidence required before a model result is reported as a detection.<\/p>\n<p>The setting uses a scale from 0 to 1:<\/p>\n<ul>\n<li><span class=\"scratchblocks\" data-scratch-input-type=\"text\">0.35<\/span> &#8211; default.<\/li>\n<li><span class=\"scratchblocks\" data-scratch-input-type=\"text\">0.5-0.7<\/span> &#8211; stricter; can reduce weak or false detections.<\/li>\n<li><span class=\"scratchblocks\" data-scratch-input-type=\"text\">0.2-0.3<\/span> &#8211; more permissive; may help with difficult detections but can also increase mistakes.<\/li>\n<\/ul>\n<p>This is a useful way to teach students that AI predictions are not simply \u201cright\u201d or \u201cwrong\u201d &#8211; each detection comes with a confidence score.<\/p>\n<p>&nbsp;<\/p>\n<div id=\"block-minimum-confidence\" class=\"scratchblocks\">minimum confidence<\/div>\n<p>Reports the current minimum confidence setting on the 0-1 scale.<\/p>\n<p><strong>Remember:<\/strong> both &#8220;minimum confidence&#8221; and &#8220;confidence of [OBJECT] #[INDEX] in [SOURCE] detections&#8221; use the same 0-1 scale.<\/p>\n<hr \/>\n<h3>\u2699\ufe0f Classification Controls<\/h3>\n<ul>\r\n<li><span class=\"scratchblocks\" id=\"block-sc-classify-interval\">classify [INTERVAL]<\/span> - Choose how often detection runs:\r\n<ul>\r\n<li>every time this block runs<\/li>\r\n<li>continuous, without delay<\/li>\r\n<li>continuous, every 50\u20132500 ms<\/li>\r\n<\/ul>\r\n<\/li>\r\n<li><span class=\"scratchblocks\" id=\"block-sc-turn-classification\">turn classification [on\/off]<\/span> - start or stop continuous detection.<\/li>\r\n<li><span class=\"scratchblocks\" id=\"block-sc-classification-interval\">classification interval<\/span> - reports the current interval in milliseconds.<\/li>\r\n<li><span class=\"scratchblocks\" id=\"block-sc-continuous-classification\">continuous classification<\/span> - reports continuous detection is \u201con\u201d or \u201coff\u201d.<\/li>\r\n<li><span class=\"scratchblocks\" id=\"block-sc-select-input-image\">select input image [camera\/stage]<\/span> - choose camera or stage.<\/li>\r\n<li><span class=\"scratchblocks\" id=\"block-sc-input-image\">input image<\/span> - reports the active input source.<\/li>\r\n<\/ul>\n<hr \/>\n<h3>\ud83c\udfa5 Video Controls<\/h3>\n<ul>\r\n    <li><span class=\"scratchblocks\" id=\"block-sc-turn-video\">turn video [off\/on\/on-flipped]<\/span>\r\n        <ul>\r\n            <li><span class=\"scratchblocks\" data-scratch-input-type=\"round\">on<\/span>: shows the camera preview in a mirrored view (like a typical webcam or mirror).<\/li>\r\n            <li><span class=\"scratchblocks\" data-scratch-input-type=\"round\">on-flipped<\/span>: shows the camera preview in a non-mirrored view \u2014 directions appear as in the real world.<\/li>\r\n            <li><span class=\"scratchblocks\" data-scratch-input-type=\"round\">off<\/span>: turns off the camera preview. In stage input mode, detection continues to run.<\/li>\r\n        <\/ul>\r\n    <\/li>\r\n    <li><span class=\"scratchblocks\" id=\"block-sc-set-video-transparency\">set video transparency to [TRANSPARENCY|text]<\/span> \u2014 adjusts how visible the camera preview is:\t \r\n        <ul>\r\n            <li><span class=\"scratchblocks\" data-scratch-input-type=\"text\">0<\/span>: fully visible (solid image)<\/li>\r\n            <li><span class=\"scratchblocks\" data-scratch-input-type=\"text\">100<\/span>: fully transparent (invisible but active)<\/li>\r\n        <\/ul>\r\n        Use a value from 0\u2013100 to adjust the camera background\u2019s opacity and help sprites appear clearly over the camera preview.<\/li>\r\n    <li><span class=\"scratchblocks\" id=\"block-sc-select-camera\">select camera [CAMERA]<\/span> \u2014 chooses among available cameras on your device. The dropdown lists all detected cameras, and the extension switches automatically to the one you select.<\/li>\r\n<\/ul>\n<ul>\r\n<li><div id=\"block-show-detection-boxes\" class=\"scratchblocks\">show detection boxes [on\/off]<\/div>\r\nTurn the detection box overlay on or off.\r\nWhen it is on, boxes appear around each detected object so you can see what the model is tracking.\r\nThis is useful for testing the camera, checking how well the model is working, and debugging a project in real time.<\/li>\r\n<\/ul>\n<p>Detection boxes are enabled by default for this extension. Use the detection-box control when you want to show or hide the visual boxes around recognized objects.<\/p>\n<hr \/>\n<h2>\ud83c\udf93 Educational Uses<\/h2>\n<ul>\n<li><strong>Introduce Computer Vision:<\/strong> Students can see how an AI model turns images into object labels and locations.<\/li>\n<li><strong>Teach Confidence and Uncertainty:<\/strong> Compare predictions at different confidence thresholds and discuss why AI sometimes makes mistakes.<\/li>\n<li><strong>Explore Coordinates:<\/strong> Connect detected X\/Y positions to sprite movement and Scratch\u2019s coordinate system.<\/li>\n<li><strong>Explore Size:<\/strong> Use width and height to see how large an object appears in the image.<\/li>\n<li><strong>Practice Lists and Indexing:<\/strong> Work with the first, second, or third matching object and compare multiple detections.<\/li>\n<li><strong>Practice Sorting:<\/strong> Save detections and order them by position, size, or confidence.<\/li>\n<li><strong>Teach Events:<\/strong> Use hat blocks to show how software can react to changes in the environment.<\/li>\n<li><strong>Discuss AI Limitations:<\/strong> Compare what students see with what the model recognizes and investigate false positives, missed objects, lighting, angle, and object size.<\/li>\n<\/ul>\n<hr \/>\n<h2>\ud83c\udfae Example Projects<\/h2>\n<ul>\n<li><strong>Object Greeter:<\/strong> Make a sprite say hello when a chosen supported object appears.<\/li>\n<li><strong>Object Counter:<\/strong> Show the number of visible objects and trigger a celebration when the count reaches a target.<\/li>\n<li><strong>Follow the Target:<\/strong> Use the X\/Y center of a detected object to move a sprite toward it.<\/li>\n<li><strong>Left or Right?<\/strong> Save matching detections from left to right and use item 1 to find the leftmost object.<\/li>\n<li><strong>Biggest Object Challenge:<\/strong> Save detections from large to small and make Scratch react to the largest one.<\/li>\n<li><strong>AI Confidence Lab:<\/strong> Display confidence values and test how lighting, distance, angle, and background affect predictions.<\/li>\n<li><strong>Classroom Object Hunt:<\/strong> Give students a list of supported objects and score a point when the AI finds each one.<\/li>\n<li><strong>What\u2019s in Frame?<\/strong> Display the &#8220;list of [SOURCE] detections&#8221; reporter block so students can observe how the model describes a changing scene.<\/li>\n<\/ul>\n<p><strong>\ud83e\udde9<\/strong> <strong>Try it yourself:<\/strong> <a href=\"https:\/\/pishi.ai\/play\" target=\"_blank\" rel=\"noopener\">pishi.ai\/play<\/a><\/p>\n<p>&nbsp;<\/p>\n<hr \/>\n<h2>\ud83d\udd27 Tips and Troubleshooting<\/h2>\n<ul>\r\n<li><strong>No camera?<\/strong> <br \/>\u2022 Make sure your camera is connected and browser permission is allowed. <br \/>\u2022 If the camera is blocked, enable it in your browser\u2019s site settings and reload the page. <br \/>\u2022 When the extension loads, if no camera is detected, the input image automatically switches to the Scratch stage so you can continue using and testing the extension's features.<\/li>\r\n<li><strong>No detection?<\/strong> <br \/>\u2022 <span class=\"scratchblocks\" id=\"block-sc-continuous-classification\">continuous classification<\/span>: Use this reporter to see if classification is active. <br \/>\u2022 If it is active, improve lighting and face the camera directly. <br \/>\u2022 <span class=\"scratchblocks\" id=\"block-sc-turn-classification\">turn classification [on]<\/span>: Use this block, if classification is not active, then recheck the classification status with the above reporter. <br \/>\u2022 In camera input mode, when the camera is turned off, classification is also stopped - you must turn the video back on or switch input to stage. <br \/>\u2022 In stage input mode, the system classifies whatever is visible on the stage - backdrops, sprites, or images. You can turn off the video completely and still process stage images. <br \/>\u2022 Stage mode is slower than camera input, so lower your classification interval (e.g., 100\u2013250 ms) for smoother results using this block: <span class=\"scratchblocks\" id=\"block-sc-classify-interval\">classify [INTERVAL]<\/span> <br \/>\u2022 In stage mode, \u201cleft\u201d and \u201cright\u201d landmarks are swapped because the stage image is not mirrored - coordinate space represents a real (non-mirrored) view. <br \/>\u2022 Classification can also restart automatically when you use blocks such as: <br \/><span class=\"scratchblocks\" id=\"block-sc-turn-video\">turn video [on]<\/span> \/ <span class=\"scratchblocks\" id=\"block-sc-classify-interval\">classify [INTERVAL]<\/span> \/ <span class=\"scratchblocks\" id=\"block-sc-select-camera\">select camera [CAMERA]<\/span> \/ <span class=\"scratchblocks\" id=\"block-sc-select-input-image\">select input image [camera\/stage]<\/span>.<\/li>\r\n<li><strong>Flipped view?<\/strong> <br \/><span class=\"scratchblocks\" id=\"block-sc-turn-video\">turn video [on-flipped]<\/span>: Use this to show the camera without mirroring. \u201con\u201d mirrors like a selfie; \u201con-flipped\u201d shows real left\/right orientation.<\/li>\r\n<li><strong>Laggy or slow?<\/strong> <br \/>Use classification intervals between 100\u2013250 ms or close other browser tabs to reduce processing load.<\/li>\r\n<li><strong>WebGL2 warning?<\/strong> <br \/>Try Firefox, or a newer device that supports WebGL2 graphics acceleration.<\/li>\r\n<li><strong>Analyze stage instead of camera?<\/strong> <br \/><span class=\"scratchblocks\" id=\"block-sc-select-input-image\">select input image [stage]<\/span>: Use this to analyze the Scratch stage image instead of a live camera feed. <span>The stage can include the camera video, as well as any sprites and backdrops currently visible on the Scratch stage.<\/span>\r\n<div><strong>Note:<\/strong> Processing the Scratch stage may be slower than using the camera directly.<\/div><\/li>\r\n<\/ul>\r\n\n<h3>\ud83c\udfaf Object Detection Specific Tips<\/h3>\n<ul>\n<li><strong>No objects detected?<\/strong> Make sure classification is running, improve the lighting, move the object closer, and check that the selected model supports that object label.<\/li>\n<li><strong>The object has the wrong name?<\/strong> Each detection model uses its own learned object vocabulary. Check the &#8220;list of [current] detections&#8221; block to see the exact label it returns.<\/li>\n<li><strong>Too many false detections?<\/strong> Increase minimum confidence gradually, for example from 0.35 to 0.5.<\/li>\n<li><strong>A real object is often missed?<\/strong> Try better lighting, a clearer angle, the Balanced model, or a slightly lower confidence threshold.<\/li>\n<li><strong>Need an object that Fast or Balanced does not recognize?<\/strong> Try the Wide model, which uses a much larger class vocabulary.<\/li>\n<li><strong>Only interested in a few objects?<\/strong> Add those exact labels to my detection list, then switch &#8220;look for [MODE]&#8221; to my detection list.<\/li>\n<li><strong>My detection list mode suddenly detects nothing?<\/strong> Check that my detection list is not empty and that its names match the model labels exactly.<\/li>\n<li><strong>Item 2 or item 3 returns empty?<\/strong> That item number only works when enough matching detections currently exist.<\/li>\n<li><strong>Saved positions seem old?<\/strong> Saved detections are intentionally a snapshot. Run &#8220;save [OBJECT] from [SOURCE] detections, order by [ORDER_TYPE]&#8221; again, usually with current as the source, to refresh them.<\/li>\n<li><strong>Project is slow?<\/strong> Use the Fast model, increase the classification interval, close other heavy browser tabs, or reduce how often your Scratch scripts read detection data.<\/li>\n<li><strong>Camera is not available?<\/strong> Switch the input to stage and test with a backdrop, sprite, or imported image.<\/li>\n<\/ul>\n<hr \/>\n<h2>\ud83d\udd12 Privacy and Safety<\/h2>\n<ul>\r\n \t<li>Everything runs locally in your browser.<\/li>\r\n \t<li>No images or video are uploaded anywhere.<\/li>\r\n \t<li>Model files are loaded directly from Pishi.ai when the extension starts, so no third-party model servers are used.<\/li>\r\n \t<li>Always ask a teacher or parent before using the camera.<\/li>\r\n \t<li>Anytime, you can safely <span class=\"scratchblocks\" id=\"block-sc-turn-video\">turn video [off]<\/span>.<\/li>\r\n<\/ul>\r\n\n<hr \/>\n<h2>\ud83e\uddea Technical Info<\/h2>\n<ul>\n<li><strong>Model Type:<\/strong> Pre-trained ONNX object detection models.<\/li>\n<li><strong>Framework:<\/strong> onnxruntime-web.<\/li>\n<li><strong>Browser Execution:<\/strong> Desktop browsers prefer WebGPU when supported, with WebAssembly (WASM) fallback; Android uses WASM.<\/li>\n<li><strong>Model Options:<\/strong> Fast (320\u00d7320, 365 classes), Balanced (640\u00d7640, 365 classes), and Wide (640\u00d7640, 4,585 classes).<\/li>\n<li><strong>Starting Model:<\/strong> Fast (320\u00d7320) on every supported device.<\/li>\n<li><strong>Default Minimum Confidence:<\/strong> 0.35.<\/li>\n<li><strong>Detection:<\/strong> Multiple objects can be reported from the same frame.<\/li>\n<li><strong>Confidence Reporter:<\/strong> Returns a value from 0-1.<\/li>\n<li><strong>Coordinates:<\/strong> Detection-box centers are mapped to Scratch stage coordinates; width and height are reported in stage units.<\/li>\n<li><strong>Sources:<\/strong> Reporter blocks can read either current live detections or saved snapshot detections.<\/li>\n<li><strong>Filtering:<\/strong> Detection mode can keep all detections or only labels from the user\u2019s custom detection list.<\/li>\n<li><strong>Saved Detections:<\/strong> A snapshot can be filtered and ordered by confidence, horizontal position, vertical position, or size.<\/li>\n<li><strong>Hat Stability:<\/strong> Detection count changes must remain the same for 2 consecutive frames before object-detection hats are treated as stable.<\/li>\n<li><strong>Detection Boxes:<\/strong> Visual detection boxes are enabled by default and can be controlled from the extension.<\/li>\n<li><strong>Inputs:<\/strong> Live camera or Scratch stage image.<\/li>\n<li><strong>Processing:<\/strong> Detection runs locally in the browser after the required model files are loaded.<\/li>\n<\/ul>\n<hr \/>\n<h2>\ud83d\udd17 Related Extensions<\/h2>\n<ul>\n<li>\ud83d\udd90\ufe0f <strong>HandPose<\/strong> &#8211; detect hand landmarks and gestures.<\/li>\n<li>\ud83d\udd7a <strong>PoseNet<\/strong> &#8211; track full-body pose and movement.<\/li>\n<li>\ud83d\ude0e <strong>FaceMesh<\/strong> &#8211; detect facial landmarks and expressions.<\/li>\n<li>\ud83d\uddbc\ufe0f <strong>Image Trainer<\/strong> &#8211; train your own custom image classifier directly in Scratch.<\/li>\n<li>\ud83c\udfeb <strong>Google Teachable Machine<\/strong> &#8211; import pre-trained custom classification models.<\/li>\n<\/ul>\n<hr \/>\n<h2>\ud83d\uddbc\ufe0f Object Detection vs. Image Trainer<\/h2>\n<div class=\"pishi-table\">\n<table>\n<thead>\n<tr>\n<th>Feature<\/th>\n<th>Object Detection<\/th>\n<th>Image Trainer<\/th>\n<\/tr>\n<\/thead>\n<tbody>\n<tr>\n<td><strong>Training<\/strong><\/td>\n<td>Uses pre-trained models &#8211; no training required<\/td>\n<td>Students train their own labels with examples<\/td>\n<\/tr>\n<tr>\n<td><strong>What it answers<\/strong><\/td>\n<td>\u201cWhat objects are here, how many, and where are they?\u201d<\/td>\n<td>\u201cWhich label best matches this whole image?\u201d<\/td>\n<\/tr>\n<tr>\n<td><strong>Multiple objects at once<\/strong><\/td>\n<td>Yes<\/td>\n<td>Not as separate located objects<\/td>\n<\/tr>\n<tr>\n<td><strong>Position and size<\/strong><\/td>\n<td>Yes &#8211; X, Y, width, and height<\/td>\n<td>No object bounding-box position<\/td>\n<\/tr>\n<tr>\n<td><strong>Custom object categories<\/strong><\/td>\n<td>Limited to labels available in the selected pre-trained detection model<\/td>\n<td>Yes &#8211; students create their own labels<\/td>\n<\/tr>\n<tr>\n<td><strong>Best classroom use<\/strong><\/td>\n<td>Computer vision, object tracking, counting, coordinates, and event-driven projects<\/td>\n<td>Teaching machine learning, training data, examples, and custom visual commands<\/td>\n<\/tr>\n<\/tbody>\n<\/table>\n<\/div>\n<p class=\"pishi-table-note\">\u2194 Swipe left or right to view full table on mobile<\/p>\n<p>&nbsp;<\/p>\n<h3>\ud83d\udca1 Why Choose Object Detection?<\/h3>\n<p><strong>Choose Object Detection when your project needs to find real objects and know where they are.<\/strong><br \/>\nIt is especially useful when students want immediate results without training a model first.<\/p>\n<ul>\n<li>Recognize several objects in the same scene.<\/li>\n<li>Count objects and compare their positions or sizes.<\/li>\n<li>Build projects that react to changes in the real world.<\/li>\n<li>Explore confidence, filtering, sorting, and computer vision in a Scratch-friendly way.<\/li>\n<\/ul>\n<h3>\ud83d\udca1 When Should You Use Image Trainer Instead?<\/h3>\n<p>Use Image Trainer when you need categories that are specific to your own project &#8211; for example, your own hand gestures, classroom cards, custom commands, or objects that a pre-trained detector does not know.<br \/>\nImage Trainer lets students create the labels themselves and learn directly from their own examples.<\/p>\n<p><strong>In short:<\/strong><br \/>\nUse Object Detection to find and locate objects the AI already knows.<br \/>\nUse Image Trainer when you want to teach the AI your own visual categories.<\/p>\n<hr \/>\n<h2>\ud83d\udcda Learn More<\/h2>\n<ul>\n<li><a href=\"https:\/\/pishi.ai\/help\" target=\"_blank\" rel=\"noopener\">AI Extensions Help at pishi.ai<\/a><\/li>\n<li><a href=\"https:\/\/onnxruntime.ai\/docs\/tutorials\/web\/\" target=\"_blank\" rel=\"noopener\">ONNX Runtime Web Documentation<\/a><\/li>\n<li><a href=\"https:\/\/scratch.mit.edu\/ideas\" target=\"_blank\" rel=\"noopener\">Scratch Ideas<\/a><\/li>\n<\/ul>\n<hr \/>\n<p><script>\nwindow.scratchBlocksConfig = {\n\tblockSpecs: [\n\t\t{ id: 'block-sc-turn-classification' },\n\t\t{ id: 'block-sc-classify-interval' },\n\t\t{ id: 'block-sc-select-input-image' },\n\t\t{ id: 'block-sc-turn-video' },\n\t\t{ id: 'block-sc-set-video-transparency' },\n\t\t{ id: 'block-sc-select-camera' },\n\t\t{ id: 'block-show-detection-boxes', blockType: 'command' },\n\t\t{ id: 'block-get-list-of-detections', blockType: 'reporter' },\n\t\t{ id: 'block-get-object-count', blockType: 'reporter' },\n\t\t{ id: 'block-get-all-items-argument', blockType: 'reporter' },\n\t\t{ id: 'block-get-nameof-detection-by-index', blockType: 'reporter' },\n\t\t{ id: 'block-get-number-of-detections-by-name', blockType: 'reporter' },\n\t\t{ id: 'block-save-and-order-detections', blockType: 'command' },\n\t\t{ id: 'block-get-confidence-of', blockType: 'reporter' },\n\t\t{ id: 'block-get-x-of', blockType: 'reporter' },\n\t\t{ id: 'block-get-y-of', blockType: 'reporter' },\n\t\t{ id: 'block-get-width-of', blockType: 'reporter' },\n\t\t{ id: 'block-get-height-of', blockType: 'reporter' },\n\t\t{ id: 'block-is-anything-detected', blockType: 'boolean' },\n\t\t{ id: 'block-is-object-detected', blockType: 'boolean' },\n\t\t{ id: 'block-set-look-for-mode' },\n\t\t{ id: 'block-get-my-detection-list-item-count', blockType: 'reporter' },\n\t\t{ id: 'block-get-my-detection-list', blockType: 'reporter' },\n\t\t{ id: 'block-get-my-detection-list-item', blockType: 'reporter' },\n\t\t{ id: 'block-add-to-my-detection-list', blockType: 'command' },\n\t\t{ id: 'block-remove-from-my-detection-list', blockType: 'command' },\n\t\t{ id: 'block-clear-my-detection-list', blockType: 'command' },\n\t\t{ id: 'block-when-any-object-detected', blockType: 'hat' },\n\t\t{ id: 'block-when-object-detected', blockType: 'hat' },\n\t\t{ id: 'block-when-detections-count-changed', blockType: 'hat' },\n\t\t{ id: 'block-set-detection-model' },\n\t\t{ id: 'block-set-minimum-confidence' },\n\t\t{ id: 'block-minimum-confidence', blockType: 'reporter' },\n\t\t{ id: 'block-sc-classification-interval', blockType: 'reporter' },\n\t\t{ id: 'block-sc-continuous-classification', blockType: 'reporter' },\n\t\t{ id: 'block-sc-input-image', blockType: 'reporter' }\n\t]\n};\n<\/script><\/p>\n","protected":false},"excerpt":{"rendered":"<p>Detect, count, locate, and track everyday objects in Scratch using AI. The Pishi.ai Object Detection extension makes computer vision easy to explore through interactive games, classroom activities, and student projects.<\/p>","protected":false},"author":1,"featured_media":8769,"comment_status":"open","ping_status":"closed","template":"","meta":{"site-sidebar-layout":"default","site-content-layout":"","ast-site-content-layout":"default","site-content-style":"default","site-sidebar-style":"default","ast-global-header-display":"","ast-banner-title-visibility":"","ast-main-header-display":"","ast-hfb-above-header-display":"","ast-hfb-below-header-display":"","ast-hfb-mobile-header-display":"","site-post-title":"","ast-breadcrumbs-content":"","ast-featured-img":"","footer-sml-layout":"","ast-disable-related-posts":"","theme-transparent-header-meta":"default","adv-header-id-meta":"","stick-header-meta":"","header-above-stick-meta":"","header-main-stick-meta":"","header-below-stick-meta":"","astra-migrate-meta-layouts":"set","ast-page-background-enabled":"default","ast-page-background-meta":{"desktop":{"background-color":"var(--ast-global-color-4)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"ast-content-background-meta":{"desktop":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"tablet":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""},"mobile":{"background-color":"var(--ast-global-color-5)","background-image":"","background-repeat":"repeat","background-position":"center center","background-size":"auto","background-attachment":"scroll","background-type":"","background-media":"","overlay-type":"","overlay-color":"","overlay-opacity":"","overlay-gradient":""}},"footnotes":""},"doc_category":[44],"doc_tag":[],"class_list":["post-8103","docs","type-docs","status-publish","has-post-thumbnail","hentry","doc_category-object-detection"],"year_month":"2026-11","word_count":3392,"total_views":"518","reactions":{"happy":"0","normal":"0","sad":"0"},"author_info":{"name":"pishi.ai","author_nicename":"pishi-aipishrobot","author_url":"https:\/\/pishi.ai\/ar\/learn\/author\/pishi-aipishrobot\/"},"doc_category_info":[{"term_name":"Object Detection Extension for Scratch","term_url":"https:\/\/pishi.ai\/ar\/help-category\/object-detection\/"}],"doc_tag_info":[],"yoast_head":"<!-- This site is optimized with the Yoast SEO plugin v28.8 - https:\/\/yoast.com\/product\/yoast-seo-wordpress\/ -->\n<title>Object Detection Extension for Scratch | Pishi.ai<\/title>\n<meta name=\"description\" content=\"Use the Object Detection extension for Scratch to recognize, count, and track objects with AI. Explore detection models, custom lists, and interactive projects.\" \/>\n<meta name=\"robots\" content=\"index, follow, max-snippet:-1, max-image-preview:large, max-video-preview:-1\" \/>\n<link rel=\"canonical\" href=\"https:\/\/pishi.ai\/ar\/help\/object-detection-detect-objects-in-scratch\/\" \/>\n<meta property=\"og:locale\" content=\"ar_AR\" \/>\n<meta property=\"og:type\" content=\"article\" \/>\n<meta property=\"og:title\" content=\"Detect Objects with AI in Scratch | Pishi.ai\" \/>\n<meta property=\"og:description\" content=\"Turn your camera into an AI vision tool. Detect, count, and track everyday objects and use them to control Scratch games and interactive projects.\" \/>\n<meta property=\"og:url\" content=\"https:\/\/pishi.ai\/ar\/help\/object-detection-detect-objects-in-scratch\/\" \/>\n<meta property=\"og:site_name\" content=\"Pishi.ai\" \/>\n<meta property=\"article:modified_time\" content=\"2026-10-08T10:09:53+00:00\" \/>\n<meta property=\"og:image\" content=\"https:\/\/pishi.ai\/wp-content\/uploads\/2026\/09\/pishi-ai-object-detection-scratch.jpg\" \/>\n\t<meta property=\"og:image:width\" content=\"1024\" \/>\n\t<meta property=\"og:image:height\" content=\"1024\" \/>\n\t<meta property=\"og:image:type\" content=\"image\/jpeg\" \/>\n<meta name=\"twitter:card\" content=\"summary_large_image\" \/>\n<meta name=\"twitter:label1\" content=\"\u0648\u0642\u062a \u0627\u0644\u0642\u0631\u0627\u0621\u0629 \u0627\u0644\u0645\u064f\u0642\u062f\u0651\u0631\" \/>\n\t<meta name=\"twitter:data1\" content=\"16 \u062f\u0642\u064a\u0642\u0629\" \/>\n<script type=\"application\/ld+json\" class=\"yoast-schema-graph\">{\"@context\":\"https:\\\/\\\/schema.org\",\"@graph\":[{\"@type\":\"WebPage\",\"@id\":\"https:\\\/\\\/pishi.ai\\\/help\\\/object-detection-detect-objects-in-scratch\\\/\",\"url\":\"https:\\\/\\\/pishi.ai\\\/help\\\/object-detection-detect-objects-in-scratch\\\/\",\"name\":\"Object Detection Extension for Scratch | Pishi.ai\",\"isPartOf\":{\"@id\":\"https:\\\/\\\/pishi.ai\\\/#website\"},\"primaryImageOfPage\":{\"@id\":\"https:\\\/\\\/pishi.ai\\\/help\\\/object-detection-detect-objects-in-scratch\\\/#primaryimage\"},\"image\":{\"@id\":\"https:\\\/\\\/pishi.ai\\\/help\\\/object-detection-detect-objects-in-scratch\\\/#primaryimage\"},\"thumbnailUrl\":\"https:\\\/\\\/pishi.ai\\\/wp-content\\\/uploads\\\/2026\\\/09\\\/pishi-ai-object-detection-scratch.jpg\",\"datePublished\":\"2026-08-29T15:00:17+00:00\",\"dateModified\":\"2026-10-08T10:09:53+00:00\",\"description\":\"Use the Object Detection extension for Scratch to recognize, count, and track objects with AI. Explore detection models, custom lists, and interactive projects.\",\"breadcrumb\":{\"@id\":\"https:\\\/\\\/pishi.ai\\\/help\\\/object-detection-detect-objects-in-scratch\\\/#breadcrumb\"},\"inLanguage\":\"ar\",\"potentialAction\":[{\"@type\":\"ReadAction\",\"target\":[\"https:\\\/\\\/pishi.ai\\\/help\\\/object-detection-detect-objects-in-scratch\\\/\"]}]},{\"@type\":\"ImageObject\",\"inLanguage\":\"ar\",\"@id\":\"https:\\\/\\\/pishi.ai\\\/help\\\/object-detection-detect-objects-in-scratch\\\/#primaryimage\",\"url\":\"https:\\\/\\\/pishi.ai\\\/wp-content\\\/uploads\\\/2026\\\/09\\\/pishi-ai-object-detection-scratch.jpg\",\"contentUrl\":\"https:\\\/\\\/pishi.ai\\\/wp-content\\\/uploads\\\/2026\\\/09\\\/pishi-ai-object-detection-scratch.jpg\",\"width\":1024,\"height\":1024,\"caption\":\"Pishi points to a screen detecting and labeling an apple, bottle, cup, backpack, ball, and toy.\"},{\"@type\":\"BreadcrumbList\",\"@id\":\"https:\\\/\\\/pishi.ai\\\/help\\\/object-detection-detect-objects-in-scratch\\\/#breadcrumb\",\"itemListElement\":[{\"@type\":\"ListItem\",\"position\":1,\"name\":\"Home\",\"item\":\"https:\\\/\\\/pishi.ai\\\/\"},{\"@type\":\"ListItem\",\"position\":2,\"name\":\"Help\",\"item\":\"https:\\\/\\\/pishi.ai\\\/help\\\/\"},{\"@type\":\"ListItem\",\"position\":3,\"name\":\"Object Detection &#8211; Detect Objects in Scratch\"}]},{\"@type\":\"WebSite\",\"@id\":\"https:\\\/\\\/pishi.ai\\\/#website\",\"url\":\"https:\\\/\\\/pishi.ai\\\/\",\"name\":\"Pishi.ai\",\"description\":\"Learn, Create &amp; Explore AI-Powered Tools\",\"publisher\":{\"@id\":\"https:\\\/\\\/pishi.ai\\\/#organization\"},\"potentialAction\":[{\"@type\":\"SearchAction\",\"target\":{\"@type\":\"EntryPoint\",\"urlTemplate\":\"https:\\\/\\\/pishi.ai\\\/?s={search_term_string}\"},\"query-input\":{\"@type\":\"PropertyValueSpecification\",\"valueRequired\":true,\"valueName\":\"search_term_string\"}}],\"inLanguage\":\"ar\"},{\"@type\":\"Organization\",\"@id\":\"https:\\\/\\\/pishi.ai\\\/#organization\",\"name\":\"Pishi.ai\",\"url\":\"https:\\\/\\\/pishi.ai\\\/\",\"logo\":{\"@type\":\"ImageObject\",\"inLanguage\":\"ar\",\"@id\":\"https:\\\/\\\/pishi.ai\\\/#\\\/schema\\\/logo\\\/image\\\/\",\"url\":\"https:\\\/\\\/pishi.ai\\\/wp-content\\\/uploads\\\/2026\\\/01\\\/pishi-ai-logo-512x512-1.png\",\"contentUrl\":\"https:\\\/\\\/pishi.ai\\\/wp-content\\\/uploads\\\/2026\\\/01\\\/pishi-ai-logo-512x512-1.png\",\"width\":512,\"height\":512,\"caption\":\"Pishi.ai\"},\"image\":{\"@id\":\"https:\\\/\\\/pishi.ai\\\/#\\\/schema\\\/logo\\\/image\\\/\"}}]}<\/script>\n<!-- \/ Yoast SEO plugin. -->","yoast_head_json":{"title":"\u0625\u0636\u0627\u0641\u0629 \u0627\u0643\u062a\u0634\u0627\u0641 \u0627\u0644\u0643\u0627\u0626\u0646\u0627\u062a \u0641\u064a \u0633\u0643\u0631\u0627\u062a\u0634 | Pishi.ai","description":"\u0627\u0633\u062a\u062e\u062f\u0645 \u0625\u0636\u0627\u0641\u0629 \u0627\u0643\u062a\u0634\u0627\u0641 \u0627\u0644\u0643\u0627\u0626\u0646\u0627\u062a \u0641\u064a \u0633\u0643\u0631\u0627\u062a\u0634 \u0644\u0644\u062a\u0639\u0631\u0651\u0641 \u0639\u0644\u0649 \u0627\u0644\u0643\u0627\u0626\u0646\u0627\u062a \u0648\u0639\u062f\u0651\u0647\u0627 \u0648\u062a\u062a\u0628\u0651\u0639\u0647\u0627 \u0628\u0627\u0633\u062a\u062e\u062f\u0627\u0645 \u0627\u0644\u0630\u0643\u0627\u0621 \u0627\u0644\u0627\u0635\u0637\u0646\u0627\u0639\u064a. \u0627\u0633\u062a\u0643\u0634\u0641 \u0646\u0645\u0627\u0630\u062c \u0627\u0644\u0643\u0634\u0641 \u0648\u0627\u0644\u0642\u0648\u0627\u0626\u0645 \u0627\u0644\u0645\u062e\u0635\u0635\u0629 \u0648\u0627\u0644\u0645\u0634\u0627\u0631\u064a\u0639 \u0627\u0644\u062a\u0641\u0627\u0639\u0644\u064a\u0629.","robots":{"index":"index","follow":"follow","max-snippet":"max-snippet:-1","max-image-preview":"max-image-preview:large","max-video-preview":"max-video-preview:-1"},"canonical":"https:\/\/pishi.ai\/ar\/help\/object-detection-detect-objects-in-scratch\/","og_locale":"ar_AR","og_type":"article","og_title":"Detect Objects with AI in Scratch | Pishi.ai","og_description":"Turn your camera into an AI vision tool. Detect, count, and track everyday objects and use them to control Scratch games and interactive projects.","og_url":"https:\/\/pishi.ai\/ar\/help\/object-detection-detect-objects-in-scratch\/","og_site_name":"Pishi.ai","article_modified_time":"2026-10-08T10:09:53+00:00","og_image":[{"width":1024,"height":1024,"url":"https:\/\/pishi.ai\/wp-content\/uploads\/2026\/09\/pishi-ai-object-detection-scratch.jpg","type":"image\/jpeg"}],"twitter_card":"summary_large_image","twitter_misc":{"\u0648\u0642\u062a \u0627\u0644\u0642\u0631\u0627\u0621\u0629 \u0627\u0644\u0645\u064f\u0642\u062f\u0651\u0631":"16 \u062f\u0642\u064a\u0642\u0629"},"schema":{"@context":"https:\/\/schema.org","@graph":[{"@type":"WebPage","@id":"https:\/\/pishi.ai\/help\/object-detection-detect-objects-in-scratch\/","url":"https:\/\/pishi.ai\/help\/object-detection-detect-objects-in-scratch\/","name":"\u0625\u0636\u0627\u0641\u0629 \u0627\u0643\u062a\u0634\u0627\u0641 \u0627\u0644\u0643\u0627\u0626\u0646\u0627\u062a \u0641\u064a \u0633\u0643\u0631\u0627\u062a\u0634 | Pishi.ai","isPartOf":{"@id":"https:\/\/pishi.ai\/#website"},"primaryImageOfPage":{"@id":"https:\/\/pishi.ai\/help\/object-detection-detect-objects-in-scratch\/#primaryimage"},"image":{"@id":"https:\/\/pishi.ai\/help\/object-detection-detect-objects-in-scratch\/#primaryimage"},"thumbnailUrl":"https:\/\/pishi.ai\/wp-content\/uploads\/2026\/09\/pishi-ai-object-detection-scratch.jpg","datePublished":"2026-08-29T15:00:17+00:00","dateModified":"2026-10-08T10:09:53+00:00","description":"\u0627\u0633\u062a\u062e\u062f\u0645 \u0625\u0636\u0627\u0641\u0629 \u0627\u0643\u062a\u0634\u0627\u0641 \u0627\u0644\u0643\u0627\u0626\u0646\u0627\u062a \u0641\u064a \u0633\u0643\u0631\u0627\u062a\u0634 \u0644\u0644\u062a\u0639\u0631\u0651\u0641 \u0639\u0644\u0649 \u0627\u0644\u0643\u0627\u0626\u0646\u0627\u062a \u0648\u0639\u062f\u0651\u0647\u0627 \u0648\u062a\u062a\u0628\u0651\u0639\u0647\u0627 \u0628\u0627\u0633\u062a\u062e\u062f\u0627\u0645 \u0627\u0644\u0630\u0643\u0627\u0621 \u0627\u0644\u0627\u0635\u0637\u0646\u0627\u0639\u064a. \u0627\u0633\u062a\u0643\u0634\u0641 \u0646\u0645\u0627\u0630\u062c \u0627\u0644\u0643\u0634\u0641 \u0648\u0627\u0644\u0642\u0648\u0627\u0626\u0645 \u0627\u0644\u0645\u062e\u0635\u0635\u0629 \u0648\u0627\u0644\u0645\u0634\u0627\u0631\u064a\u0639 \u0627\u0644\u062a\u0641\u0627\u0639\u0644\u064a\u0629.","breadcrumb":{"@id":"https:\/\/pishi.ai\/help\/object-detection-detect-objects-in-scratch\/#breadcrumb"},"inLanguage":"ar","potentialAction":[{"@type":"ReadAction","target":["https:\/\/pishi.ai\/help\/object-detection-detect-objects-in-scratch\/"]}]},{"@type":"ImageObject","inLanguage":"ar","@id":"https:\/\/pishi.ai\/help\/object-detection-detect-objects-in-scratch\/#primaryimage","url":"https:\/\/pishi.ai\/wp-content\/uploads\/2026\/09\/pishi-ai-object-detection-scratch.jpg","contentUrl":"https:\/\/pishi.ai\/wp-content\/uploads\/2026\/09\/pishi-ai-object-detection-scratch.jpg","width":1024,"height":1024,"caption":"Pishi points to a screen detecting and labeling an apple, bottle, cup, backpack, ball, and toy."},{"@type":"BreadcrumbList","@id":"https:\/\/pishi.ai\/help\/object-detection-detect-objects-in-scratch\/#breadcrumb","itemListElement":[{"@type":"ListItem","position":1,"name":"Home","item":"https:\/\/pishi.ai\/"},{"@type":"ListItem","position":2,"name":"Help","item":"https:\/\/pishi.ai\/help\/"},{"@type":"ListItem","position":3,"name":"Object Detection &#8211; Detect Objects in Scratch"}]},{"@type":"WebSite","@id":"https:\/\/pishi.ai\/#website","url":"https:\/\/pishi.ai\/","name":"\u0628\u064a\u0634\u064a.\u0625\u064a.\u0622\u064a","description":"\u062a\u0639\u0644\u0645\u060c \u0627\u0628\u062a\u0643\u0631\u060c \u0648\u0627\u0633\u062a\u0643\u0634\u0641 \u0627\u0644\u0623\u062f\u0648\u0627\u062a \u0627\u0644\u0645\u062f\u0639\u0648\u0645\u0629 \u0628\u0627\u0644\u0630\u0643\u0627\u0621 \u0627\u0644\u0627\u0635\u0637\u0646\u0627\u0639\u064a","publisher":{"@id":"https:\/\/pishi.ai\/#organization"},"potentialAction":[{"@type":"SearchAction","target":{"@type":"EntryPoint","urlTemplate":"https:\/\/pishi.ai\/?s={search_term_string}"},"query-input":{"@type":"PropertyValueSpecification","valueRequired":true,"valueName":"search_term_string"}}],"inLanguage":"ar"},{"@type":"Organization","@id":"https:\/\/pishi.ai\/#organization","name":"\u0628\u064a\u0634\u064a.\u0625\u064a.\u0622\u064a","url":"https:\/\/pishi.ai\/","logo":{"@type":"ImageObject","inLanguage":"ar","@id":"https:\/\/pishi.ai\/#\/schema\/logo\/image\/","url":"https:\/\/pishi.ai\/wp-content\/uploads\/2026\/01\/pishi-ai-logo-512x512-1.png","contentUrl":"https:\/\/pishi.ai\/wp-content\/uploads\/2026\/01\/pishi-ai-logo-512x512-1.png","width":512,"height":512,"caption":"Pishi.ai"},"image":{"@id":"https:\/\/pishi.ai\/#\/schema\/logo\/image\/"}}]}},"_links":{"self":[{"href":"https:\/\/pishi.ai\/ar\/wp-json\/wp\/v2\/docs\/8103","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/pishi.ai\/ar\/wp-json\/wp\/v2\/docs"}],"about":[{"href":"https:\/\/pishi.ai\/ar\/wp-json\/wp\/v2\/types\/docs"}],"author":[{"embeddable":true,"href":"https:\/\/pishi.ai\/ar\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/pishi.ai\/ar\/wp-json\/wp\/v2\/comments?post=8103"}],"version-history":[{"count":13,"href":"https:\/\/pishi.ai\/ar\/wp-json\/wp\/v2\/docs\/8103\/revisions"}],"predecessor-version":[{"id":8909,"href":"https:\/\/pishi.ai\/ar\/wp-json\/wp\/v2\/docs\/8103\/revisions\/8909"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/pishi.ai\/ar\/wp-json\/wp\/v2\/media\/8769"}],"wp:attachment":[{"href":"https:\/\/pishi.ai\/ar\/wp-json\/wp\/v2\/media?parent=8103"}],"wp:term":[{"taxonomy":"doc_category","embeddable":true,"href":"https:\/\/pishi.ai\/ar\/wp-json\/wp\/v2\/doc_category?post=8103"},{"taxonomy":"doc_tag","embeddable":true,"href":"https:\/\/pishi.ai\/ar\/wp-json\/wp\/v2\/doc_tag?post=8103"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}