1.四川轻化工大学，人工智能四川省重点实验室，宜宾 644000; 2.成都西测防务科技有限公司，成都 610000; 3.四川轻化工大学，企业信息化与物联网测控技术四川省高校重点实验室，宜宾64400
TP391.7
企业信息化与物联网测控技术四川省高校重点实验室开放基金(2021WYY01); 人工智能四川省重点实验室(2020RZY02); 四川轻化工大学研究生课程建设项目(KA202030); 四川省科技厅项目（2020JDJQ0075,2020YFSY0027）资助
Chen Mingju
1. Sichuan University of Science and Engineering, Artificial Intelligence Key Laboratory of Sichuan Province, Yibin, 644000, China; 3. Sichuan University of Science and Engineering, Key Laboratory of Higher Education of Sichuan Province for Enterprise Informationalization and Internet of Things, Yibin 644000, China1. Sichuan University of Science and Engineering, Artificial Intelligence Key Laboratory of Sichuan Province, Yibin, 644000, China; 2. Chengdu Xice Defence Technology CO. LTD, Chengdu 610000, China; 3. Sichuan University of Science and Engineering, Key Laboratory of Higher Education of Sichuan Province for Enterprise Informationalization and Internet of Things, Yibin 644000, China
为了实现红外和可见光图像信息的良好平衡，本文利用生成对抗网络技术，提出了一种深层次多分类的生成对抗网络红外与可见光图像融合方法。该方法将主辅思想引入到生成器的梯度和强度信息提取中，并提高了生成器卷积层的深度及浅层网络信息提取能力。在鉴别器中使用多分类器同时估计可见光和红外区域的分布。经过连续的对峙学习，使融合结果中具有显著的对比度和丰富的纹理细节信息。实验获得的信息熵及香农熵值为6.86、互信息值为13.72、标准差值为34.82、结构相似性值为0.71。对比实验结果表明，在主客观评价中，本文提出的方法获得更好的红外与可见光图像融合性能。
In order to achieve a good balance between infrared and visible image information, this paper proposes a deep multi-classification method of infrared and visible image fusion based on generative adversarial network technology. In this method, the idea of principal and auxiliary is introduced into the gradient and intensity information extraction of generator, and the depth and shallow network information extraction ability of generator convolution layer is improved. In the discriminator, multiple classifiers are used to estimate the distribution of visible and infrared regions simultaneously. After continuous face-off learning, the fusion results have remarkable contrast and rich texture details. The obtained information entropy and Shannon entropy value are 6.86, mutual information value is 13.72, standard deviation value is 34.82 and structural similarity value is 0.71.The experimental results show that the proposed method achieves better performance of infrared and visible image fusion in subjective and objective evaluation.
徐慧娴,黄坤超,陈明举,熊兴中,田洋川.基于深层次多分类生成对抗网络的红外与可见光图像融合方法[J].电子测量技术,2022,45(4):126-133
复制
