Manual initial box correction method and system based on background distinguishing

Abstract

The invention discloses a manual initial box correction method and system based on background distinguishing. The method comprises the following steps: preprocessing a received target image; judging whether the target image is larger than 100 pixels or not, and if so, receiving an instruction for manually framing the rectangular frame as an initial rectangular frame; receiving and executing an instruction of enlarging the initial rectangular frame by 1.5 times, and obtaining an extraction area of a candidate initial frame; calculating a Bhattacharyya coefficient BC1 of the foreground histogramand the background histogram of the candidate initial rectangular frame; calculating the Bhattacharyya distance between the initial rectangular frame foreground histogram and the background histogramaccording to the Bhattacharyya coefficient; correcting the manual initial rectangular frame according to the Bhattacharyya distance; and if not, receiving and executing a morphological processing instruction, and obtaining a corrected initial rectangular frame. The accuracy of the initial area can be improved, the deviation between the rectangular frame and the real target rectangular frame during initialization of the tracking algorithm is reduced, the algorithm accuracy is improved, and the target tracking accuracy is further improved.

Description

technical field [0001] The invention relates to the field of computer vision, in particular to a method and system for manual initial frame correction based on background distinction. Background technique [0002] How to improve the tracking efficiency, tracking accuracy and cost reduction of moving targets has become an important research direction in the field of computer vision in recent years. Generally, the ASMS algorithm is used to track and calculate the target, which directly initializes the algorithm with the manually framed area, and obtains the target template through the gray histogram extracted from the pixels in the area, and the tracking of subsequent frames is based on this template. The ASMS algorithm defaults that the input initial rectangular frame is an "accurate" target circumscribed rectangular frame, and then tracks it. However, in actual situations, because the manual framed rectangular frame is not necessarily "accurate", there is a certain deviation...

AI Technical Summary

Problems solved by technology

The ASMS algorithm defaults that the input initial rectangular frame is an "accurate" target circumscribed rectangular frame, and then tracks it. However, in actual situations, because the manual framed rectangular frame is not necessarily "accurate", there is a certain deviation from the real initial frame. Therefore, it is easy to be affected and reduce the tracking accuracy of the algorithm, that is, the algorithm is sensitive to the initial disturbance

Images