Infared and visible light sequential image feature level fusing method based on target detection

Abstract

An infrared and visible light sequence image character level amalgamation method based object detecting belongs to image amalgamation technology field. It uses area growth method to divide up each frame fountain image and automatically get background area and object area according to effective detecting method and takes dual-tree complex wavelet transform method to the fountain image list. After the transform, apply different amalgamation rules for object and background area to realize the list image amalgamation with character level. After get the wavelet amalgamation coefficient of different area, use dual-tree complex wavelet inverse transform to get amalgamation list image. It uses different amalgamation rules to keep interesting goal information as more as possible. The amalgamation list image owns good inflexibility, stability of time and coherence, and high calculating efficiency for using the application with dual-tree complex wavelet transform.

Description

technical field [0001] The present invention relates to a method in the technical field of image processing, in particular to a feature-level fusion method of infrared and visible light sequence images based on target detection. Background technique [0002] In the past decade or so, information fusion, especially image fusion, has aroused widespread interest and research boom around the world, and its application fields cover national defense, earth observation, machine vision, intelligent transportation, medical imaging, security inspection and industrial process, etc. Image fusion is about the collaborative use of multiple imaging sensors to combine multi-source dynamic image information from the same scene to form a multi-level intelligent information processing process that is more suitable for human and machine perception, or a single image for further image analysis and understanding. A single image contains a more accurate description of the dynamic scene than any mu...

AI Technical Summary

Problems solved by technology

This paper uses the discrete wavelet framework (DWF, or shift-invariant discrete wavelet transform) to fuse sequence images, but although the transform method they use has good shift invariance, the transform has a high degree of redundant information. The amount of calculation is large, and since in most cases we are only interested in the target information in the image rather than the pixels in it, and this method cannot retain the target information to the maximum extent due to the single fusion method

Images