A FPGA-based color target detection system and method

Abstract

The invention provides an FPGA-based color target detection system and method. The system includes Bayer signal receiving module; Bayer format interpolation to RGB module; R, G, B three-way color component addition module; R, G, B three-way color component left shift module; division module; color judgment and binarization Target module; the above modules are implemented in FPGA. The invention can detect the target with color characteristic information in real time on the high-resolution image, and segment it, so as to perform follow-up processing such as target tracking.

Description

A color target detection system and method based on FPGA technical field The invention belongs to the field of target real-time detection and tracking, and in particular relates to an FPGA-based color target detection system and method. Background technique Machine vision is a science that applies computer technology to simulate human and biological vision, using robots instead of human eyes to measure and judge targets. Visual tracking is an important branch of machine vision. In visual tracking, the primary job is to use the visual system for target detection. Only when the target is correctly confirmed can the target be detected and tracked. The process of target detection is the process of image analysis and processing. Choosing an appropriate image processing method can improve the target Real-time and robustness of detection. Existing real-time color target detection methods are mainly color area target detection methods based on PC and camera systems and embedded...

AI Technical Summary

Problems solved by technology

The color area target detection method based on the PC and camera system is to connect the camera to the PC, and then perform target detection in a pure software environment such as Microsoft Visual Studio, but this detection method needs to purchase an acquisition card matched with the camera, and the price Expensive, and because it is a pure software environment, the image processing speed is very slow and cannot meet the real-time requirements; the detection method based on the embedded color area target is implemented in ARM, MCU and FPGA, and the color area target based on FPGA is embedded The detection method uses the NiosII embedded CPU, and the algorithm runs slowly

Due to the constraints of PC performance and digital signal processing chip performance, these traditional methods can only complete quasi-real-time processing of 300,000-pixel images in terms of color image processing capabilities, and cannot process higher-resolution images in real time.

Images