Digital character detection method for moving target

Abstract

The invention discloses a digital character detection method for a moving target, and relates to the field of small target detection and machine vision. According to the main scheme, the method comprises the steps that 1, a first frame obtains the position and scale information of a moving target of an initial frame according to an existing moving target template; 2, extracting 31-dimensional fusion gradient histogram FHOG features as features to predict the position and the scale of a moving target in a subsequent frame; and 3, intercepting an area where the moving target is located from eachframe, detecting the position and the size of the target numeric character by adopting a scale pool traversal mode based on structural similarity (SSIM), and circulating to the last frame for detection. According to the method, the SSIM algorithm and the scale pool method are combined, the condition that the background in a certain area (namely, a moving block) around the target numeric characteris not prone to change is fully utilized, and the method has good precision and robustness for detecting the numeric character in the moving target under the complex conditions of illumination change, scale change, background interference and the like.

Description

technical field [0001] The invention relates to the field of target detection based on vision, and more specifically relates to a detection method for digital characters of a moving target. Background technique [0002] Object detection is of great significance and value in the field of computer vision research. Its applications include intelligent video surveillance, traffic sign recognition in vehicle camera images, defect detection on material surfaces, and lesion detection and recognition in medical imaging images. At the same time, target detection is also an important prerequisite for many high-level visual processing and analysis tasks, such as behavior analysis, event detection, scene semantic understanding, lesion detection, etc., all require the use of image processing and pattern recognition technology to detect the targets in the image and determine these The semantic type of the target object, and the specific area of ​​the target object in the image is marked. ...

AI Technical Summary

Problems solved by technology

The digital character detection method on the moving target proposed in this scheme solves the problem that the existing methods are not robust to the digital character detection scene in the moving target, and it is difficult to balance the detection accuracy and detection speed.

Images