Moving object detection method based on space-time frequency spectrum combined analysis

Abstract

The invention discloses a moving object detection method based on a space-time frequency spectrum combined analysis. For a video sequence, the method mainly comprises a step of a spatial spectrum analysis between single-frame image pixels and a step of a time spectrum analysis of a time sequence formed by pixel points corresponding to continuous multi-frame images. With the spatial spectrum analysis of the single-frame static image, mutation of information between image pixels can be detected, thereby obtaining a candidate image zone maybe containing a moving object; and the candidate image zone is used as the candidate zone of the moving object. The time spectrum of the time sequence formed by pixel points corresponding to continuous multi-frame images is analyzed in the candidate zone obtained preliminarily. With the analysis, time mutation of the same pixel point can be detected and the moving object contained by the candidate zone can be detected. And then detection on the moving object of the scene is realized. The method can be applied to motion detection of a moving object in a complex scene stably and reliably; and the operation efficiency is high.

Description

technical field [0001] The invention relates to a moving target detection method based on joint analysis of space-time frequency spectrum, in particular to a moving target detection method, which belongs to the technical field of computer vision. Background technique [0002] As a research hotspot in the field of computer vision, moving object detection has a wide range of applications in many fields such as intelligent video surveillance, image retrieval, and motion analysis. In recent years, with the development of motion feature extraction and target detection technology, a variety of moving target detection methods have been formed, mainly including frame difference method, background difference method, optical flow method, etc. The basic principle of the inter-frame difference method is to perform "subtraction" operation on two consecutive frames in the video image sequence, and realize the extraction of motion information and detection of moving objects through one-to-...

AI Technical Summary

Problems solved by technology

However, most of these works are carried out in the spatial and temporal domains of the image. Background noise and static targets interfere greatly with the detection of moving targets.

[0004] At present, there are few related methods to realize moving object detection through calculation in the frequency domain of images, and there is no moving object detection method using joint analysis of space-time spectrum.

Images