Design of Machine Vision Image Target Recognition System Based on Improved ORB Algorithm

With the rapid development of computer technology and the Internet era, machine vision is a new science and technology formed by the cross-integration of various disciplines such as image processing theory, advanced information science and digital signal processing methods based on computers. The application of computer in information processing has become an inevitable trend, it can better acquire and understand images, and it can quickly and effectively analyze, extract useful targets and provide decision makers with the required content. This can also improve work efficiency and accuracy to achieve a more precise and automated management level, and play a huge role in life. First of all, this paper uses the ORB algorithm for target recognition for the identification and positioning of static targets, but this algorithm still has the problem of low feature matching accuracy. Therefore, an improved ORB feature matching method based on multiple constraints is proposed; Secondly, it studies the design of the machine vision image target recognition system, including the overall function design of the system, the system process design and the system experimental verification.