低照度环境下小尺度人脸图像增强方法  

Small-scale face image enhancement method in low illumination environment

作　　者：龚勇[1] 张海民 GONG Yong;ZHANG Haimin(Department of Information Engineering,Xuancheng Vocational and Technical College,Xuancheng 242000,China;School of Computer and Software Engineering,Anhui Institute of Information Technology,Wuhu 241000,China)

机构地区：[1]宣城职业技术学院信息工程系,宣城242000 [2]安徽信息工程学院计算机与软件工程学院,芜湖241000

出　　处：《青岛理工大学学报》2023年第1期141-146,160,共7页Journal of Qingdao University of Technology

基　　金：安徽省教育厅自然科学重点研究项目(KJ2021A1206)。

摘　　要：为有效提升低照度人脸图像清晰度,使其更好地应用于多种领域,提出低照度环境下小尺度人脸图像增强方法。应用小波收缩去噪方法对低照度人脸图像执行去噪操作,根据韦伯费西纳定理,将去噪后低照度人脸图像划分为低对比度区、低照度区、中照度区以及饱和区4个照度区域。采用多尺度增强方法对划分后各照度区图像进行多尺度分解,增强各个尺度图像。经4个照度人脸图像融合处理,实现低照度环境下小尺度人脸图像增强。实验结果表明:该方法可实现低照度环境下人脸图像增强,图像亮度与清晰度得到显著提升,当信噪比为60 dB时,清晰度相比传统方法提高了13%左右。In order to effectively improve the clarity of the face image in low illumination environment and make it better used in many fields, a small-scale face image enhancement method in low illumination environment was proposed in this study. The wavelet shrinkage denoising method was applied to denoise the face image in low illumination. According to Weber Fischner’s theorem, the denoised face image in low illumination was divided into four illuminance regions: low contrast region, low illuminance region, medium illuminance region and saturation region. The multi-scale enhancement method was used to decompose the images of the divided illuminance regions and enhance the image of each scale. After the synthesized processing of the face images of four illuminance regions, the small-scale face image enhancement in low illumination environment was realized. The experimental results showed that this method can achieve the enhancement of face image in low illumination environment. The brightness and clarity of the enhanced image were significantly improved. When the signal-to-noise ratio was 60 dB, the clarity of the enhanced face image in low illumination was improved by about 13%, compared with traditional methods.

关 键 词：低照度环境 小尺度 人脸图像 图像增强 小波收缩 照度划分 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]