High-speed dynamic QR code identification method based on improved SURF composite algorithm

Abstract

The invention belongs to the field of dynamic detection and QR code recognition, and provides a high-speed dynamic QR code identification method based on an improved SURF composite algorithm. The high-speed dynamic QR code identification method includes the steps: after images are acquired, carrying out preliminary matching and motion compensation by using a phase correlation algorithm; carrying out matching verification on the image with good related performance by adopting a Harris feature point detection method in combination with an SURF feature descriptor, constructing a motion model by utilizing a six-parameter affine motion parameter model, and completing motion estimation and compensation to completely extract a motion target; after a target image is preprocessed, using an improvedCanny algorithm to combine with Hough straight line detection to determine the vertex of a graph, correcting image distortion, performing image smoothing by using median filtering, and using a four-threshold method to accurately divide edge information; and finally, identifying a QR code. The high-speed dynamic QR code identification method has the advantages that the moving target detection precision is high; the operation speed is high; and the recognition influence caused by various working conditions can be overcome.

Description

technical field [0001] The invention relates to the fields of dynamic detection and QR code recognition, in particular to a method for moving object detection and two-dimensional code recognition and decoding under the background of high-speed motion. Background technique [0002] Moving object detection is widely used in video surveillance, robot navigation and other fields. Moving object detection in dynamic background aims to eliminate background motion through motion estimation and compensation technology, and transform the problem into moving object detection in static background. The most commonly used motion compensation method is to use the feature matching algorithm to determine the matching feature points of images in consecutive frames, and perform motion compensation on the calculated motion vectors, but the motion estimation under this method is prone to produce more error point pairs during matching. , and it takes a lot of time, so it is not suitable for occas...

AI Technical Summary

Problems solved by technology

[0004] The invention provides a high-speed dynamic two-dimensional code recognition method based on the improved SURF composite algorithm, which can effectively improve the problem of poor timeliness when AGV detects and recognizes two-dimensional codes during high-speed movement: including detection of foreground points and background points when the camera is moving Confusion, low real-time performance of motion compensation, and high-speed motion lead to the distortion of the collected QR code image and make it difficult to recognize and read

Images