Intelligent article identification system and method based on RK3568 development board

Abstract

The invention discloses an intelligent article recognition system and method based on an RK3568 development board, and relates to the technical field of image recognition, the system is composed of the RK3568 development board, a Ubuntu system, an OpenCV library and a YOLO algorithm, the Ubuntu system is programmed on the RK3568 development board, and a dependency library required by the OpenCV library and the YOLO algorithm is installed on the RK3568 development board. According to the intelligent article identification system and method based on the RK3568 development board, the RK3568 development board is used as a hardware platform, the cost is low, large-scale deployment is facilitated, and the cost is lower; the YOLO algorithm is adopted for article recognition, the method has the advantages of being efficient and accurate, the application scene with the high real-time requirement can be met, and efficiency is higher; the system is developed based on a Ubuntu system and an OpenCV library, has good cross-platform performance and portability, is easy to deploy and maintain, and is easier to deploy; the system can realize identification and monitoring of more articles by adding more hardware equipment such as cameras and sensors, and has stronger expandability.

Description

Technical Field

[0001] The present invention relates to the technical field of image recognition, and in particular to an intelligent object recognition system and method based on an RK3568 development board. Background Art

[0002] With the continuous development of artificial intelligence technology, object recognition technology has been widely used in various fields. Object recognition technology is an application field of image recognition. It identifies and classifies different objects by analyzing the features of objects in images. Object recognition technology has shown great value in many fields due to its powerful functionality and wide application potential. For example, smart home can realize intelligent home management by identifying home objects, such as automatically adjusting air conditioning temperature and light brightness; smart security, combined with face recognition technology, can realize intelligent upgrade of home security and effectively prevent intrusion incidents. In surveillance cameras, intruders, abnormal behaviors or suspicious objects can be identified. In the retail industry, in the automated cash register system, the efficiency can be improved by identifying goods for quick settlement. In addition, it can also be used for commodity inventory management and theft prevention; logistics and warehousing management can realize automated cargo classification, sorting and tracking, and improve logistics efficiency.

[0003] Traditional object recognition methods often rely on high-performance computers and complex algorithms, which are costly and inconvenient to deploy. To this end, this application proposes an intelligent object recognition system and method based on the RK3568 development board to solve the above problems. Summary of the invention

[0004] 1. Technical issues to be solved

[0005] In view of the deficiencies in the prior art, the present invention provides an intelligent object identification system and method based on the RK3568 development board, which solves the technical problems raised in the above background.

[0006] (II) Technical solution

[0007] To achieve the above purpose, the technical solution adopted by the present invention is: an intelligent object recognition system based on RK3568 development board, the system is composed of RK3568 development board, Ubuntu system, OpenCV library and YOLO algorithm, and the Ubuntu system is burned on the RK3568 development board, and the dependent libraries required by the OpenCV library and the YOLO algorithm are installed. The system is developed based on the Ubuntu system and the OpenCV library, has good cross-platform and portability, is easy to deploy and maintain, and is easier to deploy. The system can realize the recognition and monitoring of more objects by adding more hardware devices such as cameras and sensors, and has stronger scalability;

[0008] The RK3568 development board is used as the core hardware platform of the system. The RK3568 development board is used as the hardware platform, which is low-cost and suitable for large-scale deployment with lower cost;

[0009] The Ubuntu system runs on the RK3568 development board to provide a stable operating system environment;

[0010] The OpenCV library is used for image processing and analysis;

[0011] The YOLO algorithm is used to detect objects in an image. The YOLO algorithm is used to identify objects, which is efficient and accurate, can meet application scenarios with high real-time requirements, and is more efficient.

[0012] Preferably, the YOLO algorithm can be optimized, and the optimization includes adjusting model parameters and using a more efficient network structure.

[0013] Preferably, the RK3568 development board has a hardware acceleration function, which can perform GPU acceleration and NPU acceleration to improve the operating efficiency of the system.

[0014] Preferably, the system may use multi-thread processing technology to improve the concurrent processing capability of the system.

[0015] Preferably, the RK3568 development board is connected to hardware devices such as a camera and a display device.

[0016] Preferably, the OpenCV library provides image processing functions and algorithms.

[0017] An identification method of an intelligent object identification system based on an RK3568 development board comprises the following steps:

[0018] Step 1: Image acquisition: Obtain the image of the object to be identified through a camera or other image acquisition device;

[0019] Step 2: Image preprocessing: The OpenCV library preprocesses the acquired image, including grayscale, filtering, enhancement and other operations;

[0020] Step 3: Target detection: Use the YOLO algorithm to perform target detection on the preprocessed image and identify the objects in the image;

[0021] Step 4: Result output: Output the recognition results to a display device or other application.

[0022] Preferably, the image format in step 2 is JPEG, PNG, or BMP.

[0023] (III) Beneficial effects

[0024] The beneficial effects of the present invention are:

[0025] The intelligent object recognition system and method based on the RK3568 development board uses the RK3568 development board as a hardware platform, which is low in cost and suitable for large-scale deployment at a lower cost. The YOLO algorithm is used for object recognition, which is efficient and accurate, can meet application scenarios with high real-time requirements, and is more efficient. The system is developed based on the Ubuntu system and the OpenCV library, has good cross-platform and portability, is easy to deploy and maintain, and is easier to deploy. The system can realize the recognition and monitoring of more objects by adding more hardware devices such as cameras and sensors, and has stronger scalability. BRIEF DESCRIPTION OF THE DRAWINGS

[0026] Figure 1 This is a flow chart of the identification method of the intelligent object identification system of the present invention. DETAILED DESCRIPTION

[0027] The technical solutions in the embodiments of the present invention are described clearly and completely below. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of them. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0028] The present invention provides a technical solution: an intelligent object recognition system based on an RK3568 development board, the system is composed of an RK3568 development board, an Ubuntu system, an OpenCV library and a YOLO algorithm, and the Ubuntu system is burned on the RK3568 development board, and the OpenCV library and the dependent library required by the YOLO algorithm are installed, the system can adopt multi-thread processing technology to improve the concurrent processing capability of the system, and the system can realize the recognition and monitoring of more objects by adding more hardware devices such as cameras and sensors, and the scalability is stronger;

[0029] RK3568 development board is the core hardware platform of the system. It has hardware acceleration function, which can perform GPU acceleration and NPU acceleration to improve the operation efficiency of the system. Using RK3568 development board as a hardware platform is low-cost and suitable for large-scale deployment with lower cost.

[0030] The Ubuntu system runs on the RK3568 development board to provide a stable operating system environment. The RK3568 development board is connected to hardware devices such as cameras and display devices;

[0031] The OpenCV library is used for image processing and analysis. The OpenCV library provides image processing functions and algorithms. The system is developed based on the Ubuntu system and the OpenCV library. It has good cross-platform and portability, is easy to deploy and maintain, and is easier to deploy;

[0032] The YOLO algorithm is used to detect objects in images. The YOLO algorithm can be optimized, including adjusting model parameters and using a more efficient network structure. The YOLO algorithm is used for object recognition. It is efficient and accurate, can meet application scenarios with high real-time requirements, and is more efficient.

[0033] Combination Figure 1 , an identification method of an intelligent object identification system based on RK3568 development board, comprising the following steps:

[0034] Step 1: Image acquisition: Obtain the image of the object to be identified through a camera or other image acquisition device;

[0035] Step 2: Image preprocessing: The OpenCV library preprocesses the collected images, including grayscale, filtering, enhancement and other operations. The image formats are JPEG, PNG, and BMP. OpenCV can also provide functions for detecting and describing features in images, as well as matching features between different images. OpenCV's various feature detection and matching functions include SIFT, SURF, and ORB, which assist in image preprocessing.

[0036] Step 3: Target detection: Use the YOLO algorithm to perform target detection on the preprocessed image and identify the objects in the image;

[0037] Step 4: Result output: Output the recognition results to a display device or other application.

[0038] The intelligent object recognition system can also optimize the recognition method, specifically:

[0039] Model optimization can be performed: the YOLO algorithm can be optimized to improve its detection speed and accuracy;

[0040] Hardware acceleration is available: the hardware acceleration function of the RK3568 development board, such as GPU acceleration, NPU acceleration, etc., can be used to improve the operating efficiency of the system;

[0041] Multi-threaded processing is possible: multi-threaded processing technology is used to improve the system's concurrent processing capabilities.

[0042] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the appended claims and their equivalents.

Claims

1. An intelligent object recognition system based on RK3568 development board, characterized by: The system consists of RK3568 development board, Ubuntu system, OpenCV library and YOLO algorithm. The Ubuntu system is burned on the RK3568 development board, and the dependent libraries required by OpenCV library and YOLO algorithm are installed. The RK3568 development board serves as the core hardware platform of the system; The Ubuntu system runs on the RK3568 development board to provide a stable operating system environment; The OpenCV library is used for image processing and analysis; The YOLO algorithm is used to detect objects in an image.

2. According to claim 1, an intelligent object identification system based on the RK3568 development board is characterized in that: The YOLO algorithm can be optimized, and the optimization includes adjusting model parameters and using a more efficient network structure.

3. The intelligent object identification system based on the RK3568 development board according to claim 1 is characterized in that: The RK3568 development board has a hardware acceleration function, which can perform GPU acceleration and NPU acceleration to improve the operating efficiency of the system.

4. The intelligent object identification system based on the RK3568 development board according to claim 1 is characterized in that: The system can use multi-thread processing technology to improve the concurrent processing capability of the system.

5. The intelligent object identification system based on the RK3568 development board according to claim 1 is characterized in that: The RK3568 development board is connected to hardware devices such as a camera and a display device.

6. The intelligent object identification system based on the RK3568 development board according to claim 1 is characterized in that: The OpenCV library provides image processing functions and algorithms.

7. An identification method of an intelligent object identification system based on RK3568 development board, characterized in that: The following steps are involved: Step 1: Image acquisition: Obtain the image of the object to be identified through a camera or other image acquisition device; Step 2: Image preprocessing: The OpenCV library preprocesses the acquired image, including grayscale, filtering, enhancement and other operations; Step 3: Target detection: Use the YOLO algorithm to perform target detection on the preprocessed image and identify the objects in the image; Step 4: Result output: Output the recognition results to a display device or other application.

8. The identification method of the intelligent object identification system based on the RK3568 development board according to claim 7 is characterized in that: The image format in step 2 is JPEG, PNG, or BMP.

Images