BDPartNet: Feature Decoupling and Reconstruction Fusion Network for Infrared and Visible Image

作     者：Xuejie Wang Jianxun Zhang Ye Tao Xiaoli Yuan Yifan Guo 

作者机构：Department of Computer Science and EngineeringChongqing University of TechnologyChongqing400054China Liangjiang Institute of Artificial IntelligenceChongqing University of TechnologyChongqing400054China 

出 版 物：《Computers, Materials & Continua》 (计算机、材料和连续体（英文）)

年 卷 期：2024年第79卷第6期

页      面：4621-4639页

核心收录：

学科分类：11[军事学] 080901[工学-物理电子学] 0809[工学-电子科学与技术（可授工学、理学学位）] 08[工学] 0710[理学-生物学] 070207[理学-光学] 081203[工学-计算机应用技术] 080401[工学-精密仪器及机械] 080203[工学-机械设计及理论] 0804[工学-仪器科学与技术] 0835[工学-软件工程] 0813[工学-建筑学] 0802[工学-机械工程] 0803[工学-光学工程] 0814[工学-土木工程] 0836[工学-生物工程] 0812[工学-计算机科学与技术（可授工学、理学学位）] 0702[理学-物理学] 

基　　金：supported in part by the National Natural Science Foundation of China(Grant No.61971078) Chongqing Education Commission Science and Technology Major Project(No.KJZD-M202301901) 

主　　题：Deep learning feature enhancement computer vision 

摘      要：While single-modal visible light images or infrared images provide limited information,infrared light captures significant thermal radiation data,whereas visible light excels in presenting detailed texture ***-bining images obtained from both modalities allows for leveraging their respective strengths and mitigating individual limitations,resulting in high-quality images with enhanced contrast and rich texture *** capabilities hold promising applications in advanced visual tasks including target detection,instance segmentation,military surveillance,pedestrian detection,among *** paper introduces a novel approach,a dual-branch decomposition fusion network based on AutoEncoder(AE),which decomposes multi-modal features into intensity and texture information for enhanced *** contrast enhancement module(CEM)and texture detail enhancement module(DEM)are devised to process the decomposed images,followed by image fusion through the *** proposed loss function ensures effective retention of key information from the source images of both *** comparisons and generalization experiments demonstrate the superior performance of our network in preserving pixel intensity distribution and retaining texture *** the qualitative results,we can see the advantages of fusion details and local *** the quantitative experiments,entropy(EN),mutual information(MI),structural similarity(SSIM)and other results have improved and exceeded the SOTA(State of the Art)model as a whole.