AR-based view recognition and detection system and method

Abstract

The invention relates to the technical field of augmented reality, in particular to an AR-based view recognition and detection system and method, and the system comprises an image collection module, an AI view recognition module, a feature extraction and analysis module, a database, and a result output and feedback module. The method has the beneficial effects that by adopting an advanced AI view recognition technology, the accuracy and speed of object recognition in an AR view are greatly improved, and a target object can be quickly and accurately recognized in a complex environment; the feature extraction and recognition capabilities of special materials, irregular shapes and shielded objects are remarkably improved, so that the AR view construction is more complete and accurate; through a real-time interaction feedback mechanism, recognition and detection effects can be continuously optimized according to user requirements, higher-quality and more efficient AR experience is provided for users, and wide application and deep development of the AR technology in multiple fields such as industrial manufacturing, cultural education, warehouses and intelligent security and protection are promoted.

Description

Technical Field

[0001] The present invention relates to the field of augmented reality technology, and specifically provides an AR view recognition and detection system and method. Background Art

[0002] At present, AR technology has gradually matured, and significant progress has also been made in image acquisition and recognition technology. In many scenarios, the existing technology can achieve basic image recognition and processing functions, providing users with a certain augmented reality experience.

[0003] However, there are still some deficiencies in the existing AR view recognition and detection technology. For example, in a complex environment, the recognition accuracy of the target object needs to be improved, and some objects with similar features are prone to misjudgment; the recognition speed cannot meet the requirements in some scenarios with high real-time requirements, resulting in problems such as stuttering and delay in the AR experience; for some objects with special materials, irregular shapes or occluded parts, the existing recognition algorithms are difficult to effectively extract key information, making the construction and display of the AR view incomplete and inaccurate, restricting the in-depth application and expansion of AR technology in more fields. Summary of the Invention

[0004] The purpose of the present invention is to provide an AR view recognition and detection system and method to solve the problems proposed in the above background art.

[0005] To achieve the above purpose, the present invention provides the following technical solution: An AR view recognition and detection system, the system includes:

[0006] An image acquisition module for real-time acquisition of image information in the AR scene;

[0007] An AI view recognition module for analyzing and recognizing target objects in the acquired images based on deep learning algorithms;

[0008] A feature extraction and analysis module for extracting key features from the recognized object images and performing in-depth analysis;

[0009] A database for storing a large amount of object feature information and model data to support the recognition and analysis process;

[0010] A result output and feedback module for presenting the recognition and detection results to the user and making feedback adjustments according to the user's interaction operations.

[0011] Preferably, the image acquisition module includes a high-definition camera and related sensors to obtain multi-angle image data containing the target object, and the resolution of the high-definition camera is not less than 18 million pixels to ensure the clarity of image acquisition.

[0012] Preferably, the AI view recognition module develops an algorithm model using deep learning frameworks such as TensorFlow or PyTorch, and trains the model with a large amount of labeled data to improve the recognition accuracy and speed of target objects in complex environments;

[0013] The feature extraction and analysis module uses edge detection and HOG feature extraction algorithms to effectively extract and analyze the key features of objects.

[0014] Preferably, the database is a stable and reliable database server for storing and managing a large amount of object feature data and model data, and the database server interacts with the AI view recognition module and the feature extraction and analysis module in real time to compare and analyze the collected features with the data in the database, improving the recognition accuracy.

[0015] Preferably, the system further includes a user interface, which is designed to be simple and intuitive, enabling users to interact with the system conveniently, view the recognition results and perform feedback operations;

[0016] The system also includes a testing and optimization module for conducting a large number of tests after the system is built, including recognition tests of various objects under different environments and lighting conditions, and optimizing and adjusting the system according to the test results until the system meets the expected performance indicators.

[0017] An AR view recognition and detection method is implemented using an AR view recognition and detection system, and the method includes the following steps:

[0018] a) Image acquisition: Real-time acquisition of image information in the AR scene through a high-performance image acquisition device, which includes a camera with a resolution of not less than 18 million pixels and related sensors, to obtain multi-angle image data containing the target object;

[0019] b) AI view recognition: Using the AI view recognition module based on deep learning algorithms to analyze the collected images and identify the target objects therein. The AI view recognition module is developed using the TensorFlow or PyTorch deep learning framework and trained with a large amount of labeled data;

[0020] c) Feature extraction and analysis: Extract key features such as shape, color, and texture from the recognized object images, and use algorithms such as edge detection and HOG feature extraction to deeply analyze these features to further determine the attributes and categories of the objects;

[0021] d) Database comparison: Comparing and analyzing the extracted features with the object feature information and model data stored in a stable and reliable database server to improve the recognition accuracy;

[0022] e) Result Output and Feedback: Present the recognition and detection results to the user in an intuitive manner, and adjust the results based on the user's interaction operations to optimize the subsequent recognition and detection processes.

[0023] Preferably, during the training process of the AI view recognition module, data augmentation techniques are adopted to enrich the training samples, improve the adaptability of the model to complex environments and different lighting conditions, and enhance the recognition accuracy and speed of target objects.

[0024] Preferably, the feature extraction and analysis module further includes feature extraction algorithms for special materials, irregularly shaped, and occluded objects. The algorithms effectively extract the key information of these objects, making the AR view construction more complete and accurate.

[0025] Preferably, the database server further includes a real-time update module for regularly updating object feature information and model data to adapt to the changing types of objects and recognition requirements.

[0026] Preferably, the method further includes system testing and optimization steps, that is, after the system is built, a large number of tests are carried out, including recognition tests of various objects under different environments and lighting conditions, and the system is optimized and adjusted according to the test results until the system reaches the expected performance indicators, including recognition accuracy, recognition speed, and system stability.

[0027] Compared with the prior art, the beneficial effects of the present invention are:

[0028] The AR view recognition and detection system and method proposed by the present invention, by adopting advanced AI view recognition technology, greatly improve the recognition accuracy and speed of objects in the AR view, and can quickly and accurately identify target objects in complex environments; significantly improve the feature extraction and recognition capabilities of special materials, irregularly shaped, and occluded objects, making the AR view construction more complete and accurate; through a real-time interaction feedback mechanism, it can continuously optimize the recognition and detection effects according to user needs, provide users with a better quality and more efficient AR experience, and contribute to the wide application and in-depth development of AR technology in multiple fields such as industrial manufacturing, culture and education, warehouses, and intelligent security. Brief Description of the Drawings

[0029] Figure 1 It is a block diagram of the system of the present invention. Detailed Embodiments

[0030] In order to clearly and completely describe the objectives, technical solutions of the present invention, and make the advantages more clearly understood, the following further details the embodiments of the present invention with reference to the accompanying drawings. It should be understood that the specific embodiments described herein are some embodiments of the present invention, rather than all embodiments, and are only used to explain the embodiments of the present invention, not to limit the embodiments of the present invention. All other embodiments obtained by those of ordinary skill in the art without creative efforts belong to the scope of protection of the present invention.

[0031] Embodiment 1, the present invention provides a technical solution: an AR view recognition and detection system, the system includes:

[0032] An image acquisition module for real-time acquisition of image information in the AR scene; including a high-definition camera and related sensors, obtaining multi-angle image data containing the target object, and the resolution of the high-definition camera is not less than 18 million pixels to ensure the clarity of image acquisition.

[0033] An AI view recognition module for analyzing and recognizing the target object based on the collected images using deep learning algorithms; developing an algorithm model using deep learning frameworks such as TensorFlow or PyTorch, and training the model with a large amount of labeled data to improve the recognition accuracy and speed of the target object in complex environments.

[0034] A feature extraction and analysis module for extracting key features from the recognized object images and conducting in-depth analysis; using edge detection and HOG feature extraction algorithms to effectively extract and analyze the key features of the object.

[0035] A database for storing a large amount of object feature information and model data to support the recognition and analysis process; a stable and reliable database server for storing and managing a large amount of object feature data and model data, and the database server interacts with the AI view recognition module and the feature extraction and analysis module in real time to compare and analyze the collected features with the data in the database, improving the recognition accuracy.

[0036] A result output and feedback module for presenting the recognition and detection results to the user and making feedback adjustments according to the user's interaction operations.

[0037] The system further includes a user interface, which is designed to be simple and intuitive, enabling users to conveniently interact with the system, view the recognition results, and perform feedback operations;

[0038] The system further includes a testing and optimization module for conducting a large number of tests after the system is built, including recognition tests of various objects under different environments and different lighting conditions, and optimizing and adjusting the system according to the test results until the system reaches the expected performance indicators.

[0039] Example 2, based on Example 1, proposes an AR view recognition and detection method, which is implemented using a system for AR view recognition and detection. The method includes the following steps:

[0040] a) Image acquisition: Real-time acquisition of image information in the AR scene is performed through a high-performance image acquisition device. The image acquisition device includes a camera with a resolution of not less than 18 million pixels and related sensors, and multi-angle image data containing the target object is obtained.

[0041] b) AI view recognition: The AI view recognition module based on deep learning algorithms is used to analyze the acquired images to identify the target objects therein. The AI view recognition module is developed using the TensorFlow or PyTorch deep learning framework and is trained using a large amount of labeled data. During the training process of the AI view recognition module, data augmentation techniques are adopted to enrich the training samples, improve the adaptability of the model to complex environments and different lighting conditions, and improve the recognition accuracy and speed of the target objects.

[0042] c) Feature extraction and analysis: Key features such as shape, color, and texture are extracted from the recognized object images, and these features are deeply analyzed using algorithms such as edge detection and HOG feature extraction to further determine the attributes and categories of the objects. It also includes feature extraction algorithms for special materials, irregularly shaped, and occluded objects. The algorithms effectively extract the key information of these objects, making the AR view construction more complete and accurate.

[0043] d) Database comparison: The extracted features are compared and analyzed with the object feature information and model data stored in a stable and reliable database server to improve the recognition accuracy. It also includes a real-time update module for regularly updating the object feature information and model data to adapt to the changing types of objects and recognition requirements.

[0044] e) Result output and feedback: The recognition and detection results are presented to the user in an intuitive manner, and the results are feedback-adjusted according to the user's interaction operations to optimize the subsequent recognition and detection process.

[0045] The method also includes system testing and optimization steps, that is, a large number of testing works are carried out after the system is built, including recognition tests of various objects in different environments and different lighting conditions, and the system is optimized and adjusted according to the test results until the system reaches the expected performance indicators, including recognition accuracy, recognition speed, and system stability.

[0046] Although embodiments of the present invention have been shown and described, it will be understood by those of ordinary skill in the art that various changes, modifications, substitutions and variations can be made to these embodiments without departing from the principles and spirit of the present invention, and the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. A system based on AR view recognition and detection, characterized in that: The system comprises: Image acquisition module, used to collect image information in AR scene in real time; AI view recognition module, which analyzes the collected images and identifies target objects based on deep learning algorithms; Feature extraction and analysis module, which extracts key features from the identified object images and conducts in-depth analysis; Databases, which store a large amount of object feature information and model data to support the recognition and analysis process; The result output and feedback module displays the recognition and detection results to the user and makes feedback adjustments based on user interaction operations.

2. The AR view recognition and detection system according to claim 1, characterized in that: The image acquisition module includes a high-definition camera and related sensors to obtain multi-angle image data containing the target object, and the resolution of the high-definition camera is not less than 18 million pixels to ensure the clarity of image acquisition.

3. The AR view recognition and detection system according to claim 1, characterized in that: The AI ​​view recognition module uses a deep learning framework such as TensorFlow or PyTorch to develop an algorithm model and uses a large amount of labeled data to train the model to improve the recognition accuracy and speed of target objects in complex environments; The feature extraction and analysis module uses edge detection and HOG feature extraction algorithms to effectively extract and analyze the key features of objects.

4. The AR view recognition and detection system according to claim 1, characterized in that: The database is a stable and reliable database server used to store and manage a large amount of object feature data and model data. The database server interacts with the AI ​​view recognition module and the feature extraction and analysis module in real time to compare and analyze the collected features with the data in the database to improve the accuracy of recognition.

5. The AR view recognition and detection system according to claim 1, characterized in that: The system also includes a user interaction interface, which is concise and intuitive in design, allowing users to easily interact with the system, view recognition results and perform feedback operations; The system also includes a testing and optimization module, which is used to perform a large amount of testing after the system is built, including recognition tests of various objects in different environments and lighting conditions, and optimizing and adjusting the system according to the test results until the system reaches the expected performance indicators.

6. A method for AR view recognition and detection, implemented by using the AR view recognition and detection system according to any one of claims 1 to 5, characterized in that: The method comprises the following steps: a) Image acquisition: The image information in the AR scene is collected in real time through a high-performance image acquisition device, which includes a camera with a resolution of no less than 18 million pixels and related sensors to obtain multi-angle image data containing the target object; b) AI view recognition: An AI view recognition module based on a deep learning algorithm is used to analyze the collected images and identify the target objects therein. The AI ​​view recognition module is developed using the TensorFlow or PyTorch deep learning framework and is trained using a large amount of annotated data; c) Feature extraction and analysis: Extract key features such as shape, color, and texture from the identified object image, and use edge detection, HOG feature extraction and other algorithms to conduct in-depth analysis of these features to further determine the attributes and categories of the object; d) Database comparison: compare and analyze the extracted features with the object feature information and model data stored in a stable and reliable database server to improve the accuracy of recognition; e) Result output and feedback: The recognition and detection results are displayed to the user in an intuitive manner, and the results are adjusted based on the user's interactive operations to optimize the subsequent recognition and detection process.

7. The method for AR view recognition and detection according to claim 6, characterized in that: During the training process, the AI ​​view recognition module uses data enhancement technology to enrich training samples, improve the model's adaptability to complex environments and different lighting conditions, and improve the recognition accuracy and speed of target objects.

8. The method for AR view recognition and detection according to claim 6, characterized in that: The feature extraction and analysis module also includes feature extraction algorithms for objects with special materials, irregular shapes and occluded objects. The algorithm effectively extracts key information of these objects, making the AR view construction more complete and accurate.

9. The method for AR view recognition and detection according to claim 6, characterized in that: The database server also includes a real-time update module for regularly updating object feature information and model data to adapt to changing object types and recognition requirements.

10. The method for AR view recognition and detection according to claim 6, characterized in that: The method also includes system testing and optimization steps, that is, after the system is built, a large amount of testing work is carried out, including recognition tests of various objects in different environments and under different lighting conditions, and the system is optimized and adjusted according to the test results until the system reaches the expected performance indicators, including recognition accuracy, recognition speed, and system stability.

Images