Self-learning image recognition system and method applied to AI graphic recognition toy

Abstract

This invention discloses a self-learning image recognition system and method for AI image recognition toys, comprising: an image recognition module for capturing an image of a region selected by the player and recognizing the corresponding image code; and an MCU module for caching learning image data sets, issuing the image code to be searched, and receiving the recognized image code from the image recognition module, comparing the received image code with the image code issued in the cached learning image data set. This invention, through a search game designed using image recognition technology, enables players to shift from passive to active learning. Simultaneously, this educational game provides early childhood education while occupying children's playtime, thus keeping them away from other less meaningful electronic games, increasing their interest in the game, enhancing their self-learning interest, and protecting their eyesight by keeping them away from the screen.

Description

Technical Field

[0001] This invention relates to the field of toy graphic recognition technology, and more specifically, to a self-learning image recognition system and method for AI graphic recognition toys. Background Technology

[0002] As image recognition technology matures and becomes more accurate, current image recognition technologies are applied to high-precision machines and testing equipment, which are relatively expensive. There is a lack of toys based on AI image recognition technology.

[0003] Currently, most educational products on the market (excluding electronic products such as tablets, mobile phones, and computers) are passive learning products. They cannot be based on whether a child knows how to learn or not. They can be played directly and made to be memorized by listening, resulting in low learning efficiency. Alternatively, learning through books or cards will gradually cause children to lose the fun of learning games in the long run. Moreover, they are usually one-on-one learning, without the competitive fun of learning. Summary of the Invention

[0004] To address the problems existing in the prior art, the purpose of this invention is to provide a self-learning image recognition system and method for AI graphic recognition toys.

[0005] To address the aforementioned technical problems, the present invention adopts the following technical solution; A self-learning image recognition system for AI-powered graphic recognition toys includes: The image recognition module is used to capture an image of the area selected by the player and recognize the corresponding graphic code of the image. The MCU module is used to cache the learning image data group, send out the graphic code to be searched, and receive the recognized graphic code from the graphic recognition module. It compares the recognized and received graphic code with the graphic code sent in the cached learning image data group to determine whether the player is searching correctly. The FLASH memory includes a graphics storage unit, a video storage unit, and a user information storage unit. The graphics storage unit stores image data for the image-finding game and the graphic code corresponding to each image. The video storage unit stores puzzle video data corresponding to the image data in the graphics storage unit. The user information storage unit stores user information and game data. The display module is used to display the graphic content captured by the player (similar to a gun scope) and game data, including accuracy and game commands. A mobile terminal, which connects to an MCU module via Bluetooth and obtains user information and data statistics from a user information storage unit through the MCU module; The graphic display module is used to display the image group of the graphic to be searched and other graphics.

[0006] As a further description of the above technical solution: It also includes a voice broadcast module, which receives playback instructions from the MCU module and determines whether the player has correctly selected the corresponding broadcast content.

[0007] As a further description of the above technical solution: The mobile terminal is one of a mobile phone, tablet, or laptop computer, and an APP is installed on the mobile terminal. The APP connects to the MCU module via Bluetooth.

[0008] As a further description of the above technical solution: The MCU module includes a trigger unit, which is used to trigger the control image recognition module to capture and crop the image of the current target area, and determine the image selected by the player.

[0009] The present invention also adopts the following technical solution: A method for a self-learning image recognition system applied to AI-powered graphic recognition toys includes the following recognition steps: Step 1: The host computer of the MCU module is turned on. The MCU module obtains the image data and its corresponding graphic code from the graphics storage unit and caches them. The graphics display module displays the image group to be found.

[0010] Step 2: The MCU module sends out the graphic code to be found and prompts the player through the voice broadcast module. The player then points the camera of the graphic recognition module at the selected image area in the graphic display module according to the prompt, and controls the graphic recognition module to capture the current image through the trigger unit. The graphic recognition module then recognizes the captured image, obtains the graphic code, and uploads it to the MCU module.

[0011] Step 3: The MCU module compares the recognized graphic code with the issued graphic code, and the following scenario occurs based on the comparison result: Scenario 1: If the match is consistent, it means that the correct answer has been found. The MCU module (2) plays a voice broadcast instruction that prompts the correct answer, and this single game ends. Scenario 2: If the error is different, it means it is incorrect. If the MCU module issues an incorrect voice broadcast command, the player will continue to search until the number of correct or incorrect searches reaches the preset number, at which point the game will start a new one.

[0012] Step 4: The MCU module will automatically calculate the player's accuracy rate for each round and announce it through the voice broadcast module (for example, if the player finds the correct answer 8 out of 10 times, the MCU will calculate an 80% accuracy rate and announce it). The MCU module will also upload the search results to the user information storage unit to compile game data.

[0013] As a further description of the above technical solution: In step one, the image group to be searched is displayed using one of the following methods: projected graphics, displayed on a monitor, or printed physical images.

[0014] As a further description of the above technical solution: The game data in step five includes image data of the accuracy rate and the error rate, and the game data is uploaded to the folder of the corresponding user in the user information storage unit.

[0015] Compared with the prior art, the advantages of this invention are: (1) This scheme, through a search game designed with low-cost graphic recognition technology, can enable players to change from passive learning to active learning, because players need to think about the problem and make the correct feedback; at the same time, playing this game can play an enlightenment and education role for children while occupying their game time, thus keeping them away from other less meaningful games.

[0016] (2) This scheme makes it easier for players to focus on learning these mistakes and improve their learning outcomes. At the same time, the accuracy rate of finding these pictures and the pictures that are prone to errors can be shared with other players. This data information can also be used to compete with other players and compare leaderboards. Through the competitive spirit among players, children’s interest in the game and their autonomy can be improved. Attached Figure Description

[0017] Figure 1 This is a schematic diagram illustrating the principle of the present invention; Figure 2 This is a schematic diagram of the principle of the FLASH memory of the present invention; Figure 3 This is a schematic diagram of the process of the present invention; Figure 4 This is a schematic diagram of the comparison and judgment process of the MCU module of the present invention.

[0018] Explanation of the labels in the diagram: 1. Image recognition module; 2. MCU module; 21. Trigger unit; 3. FLASH memory; 31. Image storage unit; 32. Video storage unit; 33. User information storage unit; 4. Display module; 5. Mobile terminal; 6. Voice broadcast module; 7. Graphic display module. Detailed Implementation

[0019] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention.

[0020] Please see Figure 1-4 The present invention provides Embodiment 1: A self-learning image recognition system for AI-powered graphic recognition toys includes: Image recognition module 1 is used to capture an image of the area selected by the player and recognize the corresponding image code; MCU module 2 is used to cache the learning image data group, send out the graphic code to be searched, and receive the recognized graphic code from the graphic recognition module 1. It compares the recognized and received graphic code with the graphic code sent in the cached learning image data group to determine whether the player is searching correctly. The FLASH memory 3 includes a graphics storage unit 31, a video storage unit 32, and a user information storage unit 33. The graphics storage unit 31 is used to store image data for the image search game and the graphic code corresponding to each image. The video storage unit 32 is used to store puzzle video data corresponding to the image data in the graphics storage unit 31. The user information storage unit 33 is used to store user information and game data. Display module 4 is used to display the graphic content captured by the player (similar to a gun scope) and game data, including accuracy and game commands. Mobile terminal 5 connects to MCU module 2 via Bluetooth and obtains user information and statistical information from user information storage unit 33 through MCU module 2; It also includes a voice broadcast module 6, which receives playback instructions from the MCU module 2 and determines whether the player has correctly selected the corresponding broadcast content. Graphics display module 7 is used to display the image group of the image you are looking for and other images.

[0021] First, by selecting the image group to be displayed by the image display module 7, the MCU module 2 issues a broadcast command for the image code to be found, which is then broadcast by the voice broadcast module 6. After hearing the command, the player finds the corresponding image and points the camera of the image recognition module 1 at the image display module 7 to take a picture of the image they have found. At this time, the image recognition module 1 will start the image recognition program to identify and compare the image code of the image to be found with the picture taken. If they are the same, it means that the image found by the player is correct. At this time, the MCU module 2 will send a correct signal and broadcast it through the voice broadcast module 6. At the same time, it will automatically issue the command for the next image to be found, and the player can continue to play the next image-finding game. If it is wrong, the MCU module 2 will send an error signal and broadcast it through the voice broadcast module 6. The player knows that the image found is wrong, and the player continues to find the correct image.

[0022] The correct announcements include praise-based phrases such as "Great job!", "Amazing!", and "You found the right one!", while the incorrect announcements include encouragement-based phrases such as "Keep it up!" and "Try again!". The biggest advantage of this AI-powered graphic puzzle product is that it transforms players from passive learners to active learners, as they need to think about the questions and make the correct responses. At the same time, playing this game can provide early education for children while taking up their playtime, thus keeping them away from other less meaningful games.

[0023] In this invention, the mobile terminal 5 is one of a mobile phone, tablet, or laptop computer, wherein the mobile terminal 5 has an APP installed, and the APP connects to the MCU module 2 via Bluetooth.

[0024] The app on mobile terminal 5 retrieves user information and game data from the FLASH storage 3, calculates the accuracy of the game, and records learning points, making it easy for children to observe their accuracy and compare their ranking with other players.

[0025] The MCU module 2 includes a trigger unit 21, which is used to trigger the image recognition module 1 to capture and crop the image of the current target area, and to determine the image selected by the player; the trigger unit 21 facilitates the control of the image captured by the image recognition module 1.

[0026] Please see Figure 1-4 Based on Example 1, the present invention also provides Example 2: A method for a self-learning image recognition system applied to AI-powered graphic recognition toys includes the following recognition steps: Step 1: The host computer of MCU module 2 is turned on. MCU module 2 obtains the image data and its corresponding graphic code from graphics storage unit 31 and caches them. Graphic display module 7 displays the image group to be found. Step 2: MCU module 2 sends out the graphic code to be found and prompts the player through voice broadcast module 6. The player points the camera of graphic recognition module 1 at the selected image area in graphic display module 7 according to the prompt, and controls graphic recognition module 1 to take a picture of the current image through trigger unit 21. Then, graphic recognition module 1 recognizes the captured image, obtains the graphic code, and uploads it to MCU module 2.

[0027] Step 3: MCU module 2 compares the recognized graphic code with the issued graphic code, and generates the following scenario based on the comparison result.

[0028] Scenario 1: If the match is correct, it means the correct answer has been found. MCU module 2 plays a voice prompt indicating the correct answer, and this single game ends, automatically entering the next game. Scenario 2: If the error is different, it means it is incorrect. If the MCU module 2 issues an incorrect voice broadcast command, the player will continue to search until the number of correct or incorrect searches reaches the preset number, and then the game will end and automatically enter the next game. Step 4: The MCU module will automatically calculate the player's accuracy rate for each round and announce it through the voice broadcast module (for example, if the player finds the correct answer 8 out of 10 times, the MCU will calculate an 80% accuracy rate and announce it). The MCU module will also upload the search results to the user information storage unit to compile game data.

[0029] If a player fails to find the correct image within the game's set number of attempts, the MCU module 2 will default to a failed image-finding task, and the system will record the failure. At the same time, it will automatically issue the command for the next image to find, allowing the player to continue playing the next image-finding game. This process continues until a stage of the game is completed. The AI ​​toy will then automatically calculate and announce the player's accuracy rate.

[0030] In step one, the display module is used to display the graphic content captured by the player (similar to a gun scope) and game data for display imaging. The game data includes accuracy and game commands. In this invention, the game data in step four includes image data of the accuracy rate and the error rate, and the game data is uploaded to the folder of the corresponding user in the user information storage unit 33.

[0031] By connecting the APP to the mobile terminal 5 via Bluetooth module, in addition to displaying the content on the screen, it can also obtain the accuracy rate and the images of mistakes stored in the user information storage unit 33. This allows players to focus on learning from these mistakes, improving the learning effect. At the same time, these image-finding accuracy rates and error-prone images can be shared with other players. This data information can also be used for competitions and leaderboards among other players, enhancing children's interest and autonomy in the game through the competitive spirit among players.

[0032] The above description is merely a preferred embodiment of the present invention; however, the scope of protection of the present invention is not limited thereto. Any equivalent substitutions or modifications made by those skilled in the art within the scope of the technology disclosed in the present invention, based on the technical solution and its improved concepts, should be covered within the scope of protection of the present invention.

Claims

1. A self-learning image recognition system applied to AI-powered graphic recognition toys, characterized in that, include: The image recognition module (1) is used to capture an image of the area selected by the player and recognize the corresponding image code. MCU module (2), the MCU module (2) is used to cache the learning image data group, send out the graphic code to be searched, and receive the recognition graphic code of the graphic recognition module (1), compare the recognized and received graphic code with the graphic code sent in the cached learning image data group to determine whether the player is searching correctly; The FLASH memory (3) includes a graphics storage unit (31), a video storage unit (32), and a user information storage unit (33). The graphics storage unit (31) is used to store image data for the image search game and the graphic code corresponding to each image. The video storage unit (32) is used to store puzzle video data corresponding to the image data in the graphics storage unit (31). The user information storage unit (33) is used to store user information and game data; Display module (4), the display module (4) is used to display the graphic content captured by the player and the game data for display imaging, wherein the game data includes accuracy and game commands; Mobile terminal (5), the mobile terminal (5) is connected to MCU module (2) via Bluetooth, and obtains user information and statistical information in user information storage unit (33) through MCU module (2); The graphic display module (7) is used to display the image group of the graphic to be found and other graphics.

2. The self-learning image recognition system for AI graphic recognition toys according to claim 1, characterized in that: It also includes a voice broadcast module (6), which receives playback instructions from the MCU module (2) and determines whether the player has correctly selected the corresponding broadcast content.

3. The self-learning image recognition system for AI graphic recognition toys according to claim 1, characterized in that: The mobile terminal (5) is one of mobile phones, tablets and laptops, and an APP is installed in the mobile terminal (5). The APP is connected to the MCU module (2) via Bluetooth.

4. The self-learning image recognition system for AI graphic recognition toys according to claim 1, characterized in that: The MCU module (2) includes a trigger unit (21), which is used to trigger the control image recognition module (1) to capture and extract the image of the current target area and determine the image selected by the player.

5. A method for applying a self-learning image recognition system to an AI graphic recognition toy according to any one of claims 1-4, characterized in that, The identification steps include the following: Step 1: The host, including the MCU module (2), is turned on. The MCU module (2) obtains the image data and its corresponding graphic code from the graphics storage unit (31) and caches them. The graphics display module (7) displays the image group to be found. Step 2: The MCU module (2) sends out the graphic code to be found and prompts the player through the voice broadcast module (6). The player points the camera of the graphic recognition module (1) at the selected image area in the graphic display module (7) according to the prompt, and controls the graphic recognition module (1) to take a picture of the current image through the trigger unit (21). Then the graphic recognition module (1) recognizes the captured image, obtains the recognition graphic code, and uploads it to the MCU module (2). Step 3: The MCU module (2) compares the recognized graphic code with the issued graphic code, and generates the following scenario based on the comparison result: Scenario 1: If the match is consistent, it means that the correct answer has been found. The MCU module (2) plays a voice broadcast instruction that prompts the correct answer, and the game ends. Scenario 2: If the difference is incorrect, the MCU module (2) will issue an incorrect voice broadcast command. The player will continue to search until the number of correct or incorrect searches reaches the preset number, at which point the game ends.

6. Step 4: The MCU module will automatically calculate the accuracy of the player in each round and broadcast it through the voice broadcast module (6). The MCU module (2) will also upload the search results to the user information storage unit (33) to collect game data.

7. The self-learning image recognition method applied to AI graphic recognition toys according to claim 5, characterized in that: In step one, the display module (4) displays the captured image and the data information statistically presented by the MCU.

8. The self-learning image recognition method applied to AI graphic recognition toys according to claim 5, characterized in that: The game data in step four includes image data of the accuracy rate and the error rate, and the game data is uploaded to the folder of the corresponding user in the user information storage unit (33).

Images