Children painting interactive game method, system and device and storage medium

Abstract

The invention provides a child drawing interactive game method, system and device and a storage medium, and aims to solve the problems that creation and interaction are separated, the image separation precision is insufficient, the response delay is high and multi-user concurrent interaction is difficult to support in the prior art. The method comprises the following steps: a scene construction stage: acquiring a hand-drawn image, performing image processing on the hand-drawn image to generate a semantic material with a transparent background, and mapping the semantic material into an interaction scene; a user interaction stage: acquiring spatial positioning data of user limbs, and judging whether an interaction trigger event occurs or not by adopting a collision detection algorithm according to the semantic material, the spatial positioning data and the interaction scene; and a feedback stage: outputting feedback information corresponding to the interaction trigger event. According to the method, natural fusion of child painting creation and limb interaction is realized, multi-user concurrent operation is supported, and the real-time performance, accuracy and interestingness of interaction are improved.

Description

Technical Field

[0001] This application relates to the field of children's education and entertainment technology, specifically to a method, system, device, and storage medium for children's interactive drawing game. Background Technology

[0002] Interactive games for children are a core vehicle for early childhood education, playing an irreplaceable role in stimulating creativity, cultivating teamwork skills, and enhancing sensory coordination. With the upgrading of family education needs and the popularization of artificial intelligence technology, the market is placing higher demands on the customizability, natural physical interaction, and multi-scenario adaptability of children's games.

[0003] While various technological forms exist in the current field of children's interactive games, there are still areas for improvement in deeply integrating children's free drawing creation with natural physical interaction. For example: 1. Traditional video games have fixed content and lack physical interaction and creative elements, allowing children to participate passively without integrating their own creations into the game; 2. Simple drawing interaction tools typically only support static display or basic click responses, making it difficult to achieve immersive physical interaction. The interaction methods are relatively simple, resulting in insufficient immersion; 3. Dedicated interactive devices have high hardware costs and limited scene adaptation. For example, some devices only support fixed-size artwork and are not compatible with A4, A3, etc. 4. Existing image processing-based interactive systems sometimes fail to achieve ideal separation accuracy and edge effects for the main subject in complex backgrounds, with jagged or burr-like edges affecting visual quality; 5. High interaction response latency: when using camera image processing for action recognition, the cumulative frame processing latency typically results in an overall response latency exceeding 100ms, exceeding the child's perception latency threshold and impacting interaction smoothness; 6. Lack of effective multi-target tracking and association algorithms makes it difficult to accurately distinguish concurrent stomping operations by multiple children, leading to insufficient interaction accuracy in multi-user scenarios.

[0004] In summary, current technologies have not yet been able to effectively and seamlessly integrate children's free drawing creation with natural physical interaction, resulting in a disconnect between creation and experience. Improvements are needed in terms of creative freedom, interactive smoothness, and multi-user adaptability. Summary of the Invention

[0005] This application provides a method, system, device, and storage medium for children's interactive drawing games, aiming to achieve a deep integration of children's free drawing creation and natural physical interaction, and to improve the real-time performance, accuracy, and fun of the interaction.

[0006] A method for children's interactive drawing game includes the following steps: Scene construction phase: Acquire hand-drawn images, process the hand-drawn images to generate semantic materials with transparent backgrounds, and map the semantic materials to the interactive scene; User interaction phase: Obtain spatial positioning data of the user's limbs, and use a collision detection algorithm to determine whether an interaction trigger event has occurred based on semantic materials, spatial positioning data, and interaction scenarios; Feedback phase: Output feedback information corresponding to the interactive trigger events.

[0007] Furthermore, the process of generating semantic materials includes: converting the hand-drawn image to the HSV color space for background separation, generating a mask for the alpha channel, and filtering the mask to optimize the edges.

[0008] Furthermore, the process of obtaining spatial positioning data of the user's limbs includes: Receives raw detection signals of the user's limbs collected by TOF radar; The original detection signal is analyzed and calculated to generate the corresponding point cloud data; The point cloud data is highly filtered to obtain the filtered point cloud data; The filtered point cloud data is processed by coordinate mapping to obtain the spatial positioning data of the user's limbs.

[0009] Furthermore, the point cloud data is highly filtered to include: excluding data corresponding to ground interference clutter and retaining point cloud data corresponding to user limbs.

[0010] Furthermore, the process of determining whether an interaction trigger event has occurred using a collision detection algorithm includes: Establish a coordinate mapping from the user's spatial positioning data to the interactive scene; Based on coordinate mapping, spatial positioning data is converted into coordinate positions in the interactive scene; Determine whether the coordinate position falls within the collision detection area corresponding to the semantic material in the interactive scene; When the coordinates fall within the collision detection area, an interaction trigger event is determined to have occurred.

[0011] Furthermore, the feedback information corresponding to the output interaction trigger event includes: Multi-channel audio and visual feedback; audio feedback adopts a directional sound design, matching the corresponding sound channel output according to the projection area of ​​the user's limb in the interactive trigger event, supporting at least two channels; visual feedback uses a 3D rendering engine to render dynamic effects in real time, including at least one of character highlighting, particle explosion or progress bar update.

[0012] A children's drawing interactive game system, comprising: The scene construction module is used to acquire hand-drawn images, process the hand-drawn images to generate semantic materials with transparent backgrounds, and map the semantic materials to the interactive scene; The user interaction module is used to obtain spatial positioning data of the user's limbs, and uses a collision detection algorithm to determine whether an interaction trigger event has occurred based on semantic materials, spatial positioning data, and interaction scenarios. The feedback module is used to output feedback information corresponding to interactive trigger events.

[0013] Furthermore, the user interaction module includes a spatial positioning data acquisition module; The spatial positioning data acquisition module is used to receive the original detection signals of the user's limbs collected by the TOF radar, analyze and calculate the original detection signals to generate corresponding point cloud data, perform height filtering on the point cloud data to obtain filtered point cloud data, and perform coordinate mapping processing on the filtered point cloud data to obtain the spatial positioning data of the user's limbs.

[0014] An interactive drawing game device for children, comprising: The scene construction module is used to acquire hand-drawn images, process the hand-drawn images to generate semantic materials with transparent backgrounds, and map the semantic materials to the interactive scene; The user interaction module is used to obtain spatial positioning data of the user's limbs, and uses a collision detection algorithm to determine whether an interaction trigger event has occurred based on semantic materials, spatial positioning data, and interaction scenarios. The feedback module is used to output feedback information corresponding to interactive trigger events.

[0015] A computer-readable storage medium having a computer program stored thereon, the computer program being executed by a processor to implement a method for interactive drawing games for children.

[0016] The technical solution provided in this application has at least the following advantages compared with the prior art: It achieves a natural integration of painting creation and physical interaction: through a three-stage closed-loop design of "scene construction - user interaction - feedback", children's hand-drawn images are converted into interactive virtual scene elements in real time, and the interactive response is directly driven by physical movements, which solves the problem of the separation between the creation process and the game experience in traditional solutions.

[0017] Improved image processing accuracy and edge quality: Employing background separation technology based on the HSV color space and combining it with filtering optimization to generate high-quality transparent channel masks, effectively improving the separation accuracy and edge smoothness of hand-drawn subjects, providing high-quality visual materials for subsequent precise interaction.

[0018] It ensures low latency and high precision in limb interaction: By using TOF radar to collect spatial signals, and through point cloud data processing, height filtering and coordinate mapping, it achieves millimeter-level and millisecond-level real-time positioning of user limbs (such as feet), effectively eliminating environmental interference and providing a reliable data foundation for real-time interaction.

[0019] Effectively supports concurrent interaction scenarios for multiple users: Based on high-precision spatial positioning data and collision detection algorithms for virtual scene elements, combined with spatiotemporal correlation analysis of multiple limb touch points, it can accurately distinguish and respond to concurrent operations of different users, enhancing the system's collaboration and scenario adaptability.

[0020] It provides immersive multi-channel interactive feedback: by combining directional sound effects that match the interaction location with real-time visual dynamic effects generated by a 3D rendering engine, it creates a multi-sensory feedback experience with a strong sense of space and high fun, significantly enhancing children's sense of participation and immersion.

[0021] A modular and easily expandable system architecture has been formed: the unified technical system of the method, system, device and storage medium proposed in this application has built a complete solution from software algorithm to hardware implementation. The modular design supports flexible configuration and is easy to adapt to different application scenarios from home to educational institutions and to expand functions. Attached Figure Description

[0022] Figure 1 A flowchart illustrating a children's interactive drawing game method provided in this application; Figure 2 This application provides a schematic diagram of the components of a children's interactive drawing game system. Figure 3 A schematic diagram illustrating the spatial positioning and collision detection principle of an interactive children's drawing game provided in this application; Figure 4 This is a schematic diagram of the components of a children's interactive drawing game device provided in this application. Detailed Implementation

[0023] The specific embodiments of this application will be described in detail below with reference to the accompanying drawings, so that those skilled in the art can understand and implement the children's drawing interactive game method, system, device and storage medium claimed in this application.

[0024] The core inventive concept of this application lies in converting children's hand-drawn images into interactive semantic materials in real time through a closed-loop process of scene construction, user interaction, and feedback, and achieving low-latency, immersive physical interaction based on high-precision spatial positioning and collision detection. The following embodiments are further specific descriptions of the technical solutions in the claims and should not be construed as limiting the scope of protection of this application.

[0025] This embodiment details the children's drawing interactive game method provided in this application. The method mainly includes a scene construction stage, a user interaction stage, and a feedback stage, realizing a complete closed loop from drawing input to interactive feedback.

[0026] like Figure 1 As shown, the method specifically includes the following steps: During the initial scene construction phase, the system optionally acquires interaction flow configuration information. This configuration information can be preset by the system or adjusted by the user as needed, and may include drawing acquisition resolution, background color recognition mode, collision detection threshold, feedback sound effect type, etc. The configuration information is associated with the user session and is used to guide and record the scene construction process, ensuring the controllability and reproducibility of the fully automated execution.

[0027] Next, image acquisition and semantic material generation are performed. The system prompts the user to upload background and character drawings sequentially through the user interface. The image acquisition module guides the user to scan or photograph their hand-drawn background and character artworks for non-contact 2D image capture. The image processing flow then proceeds: the acquired images are converted to the HSV color space. The HSV space more easily separates color and brightness information, accurately identifying preset or adaptively detected background colors and generating an Alpha transparency channel mask (Alpha mask). The Alpha mask is then filtered to optimize edges and reduce jagged edges. The semantic material generated in this process performs well in terms of subject segmentation accuracy, with an average Intersection over Union (IoU) value of no less than 0.85 with the actual labeled regions.

[0028] Next, scene integration and deployment are performed. The generated semantic materials, their associated interaction trigger conditions, and corresponding feedback rules are packaged and uploaded to the cloud or local game scene management platform via a predefined application programming interface. Upon receiving the data, the platform automatically registers the new materials, overlays scene element layers, and binds the interaction logic. Without requiring additional user configuration, the interactive game can be deployed and updated online in real time, allowing users to directly enter the updated scene for interaction.

[0029] User interaction phase First, spatial positioning data of the user's limbs is collected, taking the foot as an example, such as... Figure 3As shown, the electronic device, equipped with TOF radar, image acquisition equipment, and other hardware, works in tandem to acquire spatial positioning data of the user's limbs in the deployed game scene. Based on the time-of-flight ranging principle, TOF radar effectively filters ambient light interference, supports low-latency raw detection signal transmission and multi-target recognition, and boasts a spatial resolution down to the millimeter level, accurately capturing information such as the position and speed of a child's feet. The processor processes the raw detection signals to generate point cloud data. The point cloud data processing flow includes: Z-axis height filtering, based on children's height and interaction habits data, retaining point cloud data within the range of 2 cm to 30 cm above the ground. This range is suitable for the foot height of children aged 3-12 years. When the height is below 2 cm, ground debris is filtered out, and when the height is above 30 cm, non-foot objects are filtered out; coordinate mapping, converting physical space coordinates (x, y, z) into screen pixel coordinates (u, v) using the formulas u = x × screen width / physical detection range and v = y × screen height / physical detection range, realizing the correspondence between physical stepping position and scene coordinates; spatial positioning data association analysis of user limbs, determining the association relationship between two limb touch points based on Euclidean distance and a preset 100ms time window. This time window is set based on the average interval of children's continuous actions. Two adjacent limb touch points within 100ms are regarded as the feet of the same user, realizing multi-user differentiation. The processed positioning data is output in a structured format, including parameters such as the user's limb spatial positioning data ID, real-time coordinates, and movement speed, and is synchronized to the interactive judgment module to provide data support for subsequent trigger judgments. After optimization, the total delay from signal acquisition to coordinate output in this step does not exceed 20ms, which helps to solve the problem of high response latency in traditional systems.

[0030] Based on the acquired location data, interaction trigger determination is performed. The interaction determination module combines semantic materials with foot spatial positioning data and employs a specific collision detection algorithm, such as the axis-aligned bounding box (AABB) collision detection algorithm, to determine the interaction trigger. The axis-aligned bounding box algorithm establishes a bounding box (collision detection region) for each semantic material in scene coordinates. It achieves fast collision detection by determining whether the limb touch point coordinates fall within this collision detection region. The principle can be referenced elsewhere. Figure 3 First, semantic materials are layered according to the game scene coordinate system to establish a mapping relationship between physical space and scene. Then, the pixel coordinates of limb touch points and the coordinate areas of character materials are compared in real time, while scene background materials are not involved in collision detection. To distinguish different users, a threshold of at least 50 centimeters between limb touch points is set to identify different users. This threshold is determined based on statistics of the foot spacing of 200 children aged 3-12, which can cover most scenarios and supports single-person dual-limb touch point collaboration and multi-person independent operation in parallel response. When the overlap area between the limb touch point and the character material area reaches a preset threshold of more than 50%, it is determined to be a valid interaction trigger.

[0031] Based on the judgment results, multi-channel feedback output is executed. Based on the interaction-triggered judgment results, the feedback module controls the feedback device to output multi-channel feedback. In terms of sound effects, a directional sound design is adopted, matching the corresponding channel output sound effect according to the projection area of ​​the limb's contact point. It supports at least two channels; for example, stepping on the left area triggers the left channel sound effect, and stepping on the right area triggers the right channel sound effect, enhancing spatial immersion. Visually, an integrated 3D rendering engine renders dynamic effects such as character highlighting, particle explosions, and progress bar updates in real time, enhancing the interactive fun.

[0032] For example, a children's drawing interactive game system is provided. (e.g.) Figure 2 As shown, the system includes an image acquisition device, a TOF radar, a processor, and a feedback device.

[0033] An image acquisition device is used to acquire children's hand-drawn images. A Time-of-Flight (TOF) radar is used to acquire raw detection signals of the user's limbs. A processor is electrically connected to the image acquisition device and the TOF radar, and is configured to execute the children's drawing interactive game method provided in this application. A feedback device is connected to the processor and is used to output multimedia feedback information such as audio and video according to the processor's instructions.

[0034] As an entity composed of hardware modules, the system's operation involves data transfer and collaborative control between these modules. During the scene construction phase, the image acquisition device acquires hand-drawn images under the control of the processor and transmits them to the processor through a data interface; the processor executes the image processing program to generate semantic materials and integrate them into the interactive scene.

[0035] During the user interaction phase, the TOF radar continuously collects raw detection signals from the user's limbs and transmits them to the processor. The processor processes the signals in real time to obtain spatial positioning data and performs collision detection based on semantic materials in the scene to determine the interaction event.

[0036] During the feedback phase, the processor generates control commands based on the judgment results, driving the feedback device to output corresponding sound and visual effects.

[0037] In the above process, the image acquisition device, TOF radar, processor and feedback device achieve closed-loop collaboration through electrical connection and data protocol, forming a complete interactive game system for children's drawing.

[0038] During the user interaction phase, the TOF radar continuously scans the interaction area, collecting raw detection signals from the user's limbs, and streams these signals to the processor via a high-speed interface (such as SPI, UART, or Ethernet). Upon receiving the signals, the processor initiates a radar data processing thread, sequentially performing signal analysis, point cloud generation, altitude filtering, and coordinate mapping, outputting low-latency spatial positioning data.

[0039] Simultaneously, the processor's interaction determination logic is activated: it synchronously acquires semantic material information and real-time updated spatial positioning data in the scene, runs collision detection algorithms (such as AABB), determines whether an interaction event has occurred, and can distinguish between multiple users based on the distance and time relationship between touch points. The entire positioning and determination process is executed in parallel within the processor under a multi-threaded or pipelined architecture, ensuring that the overall response latency is less than 20ms.

[0040] During the feedback phase, once a valid interaction is detected, the processor immediately generates corresponding control commands, driving the feedback devices through the audio and video output interfaces respectively. For example, it controls the speakers to play the corresponding channel's audio effects, while simultaneously controlling the projector or monitor to render real-time visual effects, thereby achieving multi-channel immersive feedback with synchronized audio and video.

[0041] In summary, this system achieves closed-loop collaboration of three core data links: image acquisition device → processor, TOF radar → processor, and processor → feedback device. This enables the entire process of hardware-based operation, from image acquisition and real-time positioning to intelligent judgment and immersive feedback, forming a low-latency, high-precision, multi-user concurrent interactive game system. For example, an interactive drawing game device for children is provided. Figure 4 As shown, the device includes the following modules: The scene construction module is used to acquire hand-drawn images, perform image processing on the hand-drawn images to generate semantic materials with transparent backgrounds, and map the semantic materials to the interactive scene.

[0042] The user interaction module is used to obtain spatial positioning data of the user's limbs, and based on semantic materials, spatial positioning data and interaction scenarios, uses a collision detection algorithm to determine whether an interaction trigger event has occurred.

[0043] The feedback module is used to output feedback information corresponding to interactive trigger events.

[0044] This device can be implemented through software, hardware, or a combination of both. For example, it can be integrated into a processor to execute computer program instructions to complete the functions of each module, or it can be implemented through dedicated logic circuits. The modules communicate with each other through data interfaces or internal buses to collaboratively complete the interaction process.

[0045] For example, a computer-readable storage medium, such as a USB flash drive, hard disk, optical disk, flash memory, etc., is provided. A computer program is stored thereon, and when the computer program is loaded and executed by a processor, it can implement all the steps of the children's drawing interactive game method as described in detail in Embodiment 1.

[0046] This embodiment achieves seamless integration of children's custom creation and physical interaction through the deep fusion of drawing scanning and stomping interaction, lowering the barrier to game creation. By employing standardized process control, high-precision visual analysis technology, low-latency spatial positioning solutions, and intelligent interactive logic design, a children's interactive game system adaptable to multiple users and scenarios is constructed, suitable for kindergarten early childhood education, family parent-child interaction, and other similar scenarios.

[0047] For example, an industrial-grade Time-of-Flight (TOF) radar is selected. This module integrates a multi-zone TOF ranging chip and a high-speed data transmission interface. Through hardware-level configuration optimization, it achieves raw detection signal transmission with a latency of no more than 10ms and a resolution of 1mm. The specific technical solution is as follows: High-precision point cloud data resolution is achieved by: the radar has a built-in laser light source and uses pulse modulation technology to optimize spatial accuracy by improving the temporal resolution of the laser pulse; at the same time, it is equipped with a pixel array receiver, and the ranging accuracy of each pixel is independently calibrated. The calibration is performed through a standard calibration board, and finally, high-precision point cloud data resolution is achieved in the entire field of view to meet the positioning needs of children's subtle movements.

[0048] Low-latency data transmission is achieved through high-speed data interfaces (such as SPI, UART, or Ethernet interfaces) between the radar and the processor, which improves the transmission rate by 8 times compared to the traditional I2C protocol. At the same time, the radar's hardware data stream compression function is enabled, and the data compression ratio can reach 3:1. The processor is equipped with a dedicated DMA (Direct Memory Access) channel, so the original radar detection signal is directly written to the memory without passing through the processor core, reducing intermediate links in data transmission. According to actual tests, the total latency from the radar acquiring the original detection signal to the processor receiving it is stable at 7-9ms.

[0049] For example, a consumer-grade TOF radar is selected, and software algorithm optimization is used to compensate for hardware specification shortcomings, achieving a raw detection signal transmission latency of no more than 10ms and a resolution of 1mm. This is suitable for cost-sensitive application scenarios. The specific technical solution is as follows: High-precision point cloud data resolution is achieved by using a multi-frame data fusion algorithm, with the radar acquiring the original detection signal at a frequency of 30fps; the data of multiple consecutive frames (e.g., 3 frames) is fused using a neighborhood averaging filtering algorithm to remove random noise interference, and a sub-pixel interpolation algorithm is used to improve the effective ranging resolution to 1 mm; the fusion and interpolation algorithms are accelerated by the processor's GPU.

[0050] Low-latency data transmission is achieved by using the UDP protocol for data transmission between the radar and the processor and optimizing the data frame structure. At the same time, the radar data output strategy is optimized by adopting an incremental transmission mode. According to actual measurements, the total latency of radar data acquisition, algorithm processing, and transmission to the processor is 8-10ms.

[0051] In this embodiment, the data acquisition and processing module uses the same height filtering, coordinate mapping, and limb contact point association analysis logic as the previous TOF radar embodiment for the optimized original detection signal. After testing, the positioning accuracy is basically the same as that of the first embodiment, and it can effectively support multi-user concurrent stepping interaction.

[0052] The two embodiments described above achieve a TOF radar original detection signal transmission latency of no more than 10ms and a resolution of 1mm through hardware configuration optimization and software algorithm optimization, respectively. Those skilled in the art can choose the appropriate technical solution based on actual cost and scenario requirements, or combine the core methods of both paths for optimization, all of which can achieve the desired functionality. Furthermore, both embodiments utilize the same image processing flow and collision detection algorithm. Testing has verified that the image parsing and background separation accuracy of the proposed technical solution is effectively improved. The overall interactive response latency (including image processing, radar positioning, collision detection, and feedback output) is controlled within 20ms, meeting the real-time and accuracy requirements of children's interactive games.

[0053] For example, image processing under complex backgrounds. A high-resolution industrial area scan camera was selected as the image acquisition device. After converting the acquired RGB image to the HSV color space, specific threshold ranges for hue (H), saturation (S), and lightness (V) were set for materials commonly used in children's drawing, generating an alpha transparency channel mask. The filtering process employed joint filtering: first, Gaussian blur was performed, followed by bilateral filtering, with specific filter kernel parameters provided (e.g., Gaussian kernel variance σ = 1.2).

[0054] An example is image processing under non-uniform lighting. To address uneven illumination, grayscale equalization preprocessing is added before RGB-HSV conversion, for example, using the Limiting Contrast Adaptive Histogram Equalization (CLAHE) algorithm. After generating the alpha mask, the joint filtering parameters are adjusted, and morphological closing operations are added to fill holes in the mask.

[0055] For example, in a home use scenario, the image acquisition device uses a 16-megapixel USB area scan camera (replacing an industrial-grade camera to reduce costs), connected to the processor via a USB 3.0 interface, supporting continuous acquisition at 20fps; the TOF radar communicates with the processor via a high-speed data interface (such as SPI, UART, or Ethernet); feedback devices include a dual-channel Bluetooth speaker and a home projector (supporting 1080P resolution and a 60Hz refresh rate). The processor can run a corresponding operating system and call image processing libraries and 3D rendering engines for development. The collaborative workflow of each module is as follows: camera acquires images → image processing module generates semantic materials → scene integration module deploys to the Unity scene → TOF radar acquires positioning data → data acquisition and processing module outputs structured data → interaction judgment module performs collision detection → feedback module drives the speaker and projector to output effects. This system can be implemented with relatively low-cost hardware configurations and is suitable for home desktops and living room scenarios.

[0056] For example, in an educational institution setting, an enhanced system for that setting is implemented. The image acquisition device uses a 25-megapixel industrial camera (supporting full A3 paper coverage), connected to the processor via a GigE interface, and equipped with an autofocus lens. Simultaneously, two TOF radars are deployed to achieve full coverage of a 3m x 4m interactive area. Radar data is aggregated to the processor via high-speed data interfaces (such as SPI, UART, or Ethernet). Feedback devices include a 4-channel audio system, a 4K laser projector, and a vibration feedback pad (generating slight vibrations when stepped on). The system supports up to four children interacting concurrently. The processor uses multi-threading to ensure parallel operation of each module. The image processing module, data acquisition and processing module, and interaction determination module each occupy independent threads, ensuring smooth operation even in multi-user scenarios. This solution enhances the system's scenario scalability through multi-device adaptation and collaboration.

[0057] The above description is merely a preferred embodiment and the technical principles employed in this application. This application is not limited to the specific embodiments described herein, and various obvious changes, readjustments, and substitutions that can be made by those skilled in the art will not depart from the protection scope of this application.

Claims

1. A method for children's interactive drawing game, characterized in that, Including the following steps: Scene construction phase: Acquire hand-drawn images, process the hand-drawn images to generate semantic materials with transparent backgrounds, and map the semantic materials to the interactive scene; User interaction phase: Acquire spatial positioning data of the user's limbs, and use a collision detection algorithm to determine whether an interaction trigger event has occurred based on the semantic material, the spatial positioning data, and the interaction scenario; Feedback phase: Output the feedback information corresponding to the interactive trigger event.

2. The children's drawing interactive game method according to claim 1, characterized in that, The process of generating semantic materials includes: converting the hand-drawn image to the HSV color space for background separation, generating a mask for the transparent channel, and filtering the mask to optimize the edges.

3. The children's drawing interactive game method according to claim 1, characterized in that, The process of obtaining spatial positioning data of the user's limbs includes: Receives raw detection signals of the user's limbs collected by TOF radar; The original detection signal is analyzed and calculated to generate corresponding point cloud data; The point cloud data is highly filtered to obtain filtered point cloud data. The filtered point cloud data is processed by coordinate mapping to obtain the spatial positioning data of the user's limbs.

4. The children's drawing interactive game method according to claim 3, characterized in that, The high-level filtering of point cloud data includes: excluding data corresponding to ground interference clutter and retaining point cloud data corresponding to user limbs.

5. The children's drawing interactive game method according to claim 1, characterized in that, The process of determining whether an interactive trigger event has occurred using a collision detection algorithm includes: Establish a coordinate mapping from the user's spatial positioning data to the interactive scene; Based on the coordinate mapping, the spatial positioning data is converted into coordinate positions in the interactive scene; Determine whether the coordinate position falls within the collision detection area corresponding to the semantic material in the interactive scene; When the coordinates fall within the collision detection area, the interaction trigger event is determined to have occurred.

6. The children's drawing interactive game method according to claim 1, characterized in that, The feedback information corresponding to the interactive trigger event includes: Multi-channel audio and visual feedback; wherein, the audio feedback adopts a directional sound design, matching the corresponding sound channel output according to the projection area position of the user's limb in the interactive trigger event, supporting at least two channels; the visual feedback renders dynamic effects in real time through a 3D rendering engine, including at least one of character highlighting, particle explosion, or progress bar update.

7. A children's drawing interactive game system, characterized in that, include: The scene construction module is used to acquire hand-drawn images, process the hand-drawn images to generate semantic materials with transparent backgrounds, and map the semantic materials to the interactive scene; The user interaction module is used to obtain spatial positioning data of the user's limbs, and to determine whether an interaction trigger event has occurred based on the semantic material, the spatial positioning data, and the interaction scenario using a collision detection algorithm. The feedback module is used to output feedback information corresponding to the interactive trigger event.

8. The children's drawing interactive game system according to claim 6, characterized in that, The user interaction module includes a spatial positioning data acquisition module; The spatial positioning data acquisition module is used to receive the original detection signal of the user's limb collected by the TOF radar, analyze and calculate the original detection signal to generate corresponding point cloud data, perform height filtering on the point cloud data to obtain filtered point cloud data, and perform coordinate mapping processing on the filtered point cloud data to obtain the spatial positioning data of the user's limb.

9. A children's interactive drawing game device, characterized in that, include: The scene construction module is used to acquire hand-drawn images, process the hand-drawn images to generate semantic materials with transparent backgrounds, and map the semantic materials to the interactive scene; The user interaction module is used to obtain spatial positioning data of the user's limbs, and to determine whether an interaction trigger event has occurred based on the semantic material, the spatial positioning data, and the interaction scenario using a collision detection algorithm. The feedback module is used to output feedback information corresponding to the interactive trigger event.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that, The computer program is executed by a processor to implement the children's drawing interactive game method according to any one of claims 1-5.

Images