基于偏振图像融合的长波红外人脸图像增强技术  

LWIR Human Face Image Enhancement Method Based on Polarimetric Image Fusion

作　　者：蔡怀宇[1,2] 杨建乔[1,2] 黄战华[1,2] 朱攀[1,2] 

机构地区：[1]天津大学精密仪器与光电子工程学院,天津300072 [2]光电信息技术教育部重点实验室,天津300072

出　　处：《纳米技术与精密工程》2016年第4期262-268,共7页Nanotechnology and Precision Engineering

摘　　要：针对非制冷型长波红外相机人脸成像模糊的缺点,提出一种基于偏振的红外图像增强方法,该方法在支持度变换基础上有效地将人脸红外偏振信息和红外强度信息相融合.搭建分时偏振成像系统获取红外偏振图像,求解得到红外强度和红外偏振度图像;对它们使用支持度变换方法进行多尺度分解后,采取相应的融合规则进行融合得到最终融合增强图像.实验结果从主观人眼评价和客观参数评价两方面表明,增强图像拥有比红外图像更加丰富的人脸轮廓信息.Since the face image obtained by the uncooled long-wavelength infrared （LW1R） camera is blurred, an infrared image enhancement method based on polarization is proposed. The fusion of infrared polarization and infrared intensity human face images based on support value transform method can effec- tively improve the contrast of infrared human face image and increase details. Time-division polarimetric imaging system was constructed to obtain infrared polarimetric images and calculate infrared intensity image and infrared polarization image. These images were decomposed into multi-scale images using support value transform method, then the decomposed images were combined in the same scale together using correspond- ing fusion rules. Both subjective and objective evaluation results of the experiments show that the enhanced image obtained by the proposed method has more detailed information compared with the infrared image.

关 键 词：图像增强 人脸红外图像 红外偏振成像 支持度变换 图像融合 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]