The Study of Image Enhancement Techniques by Applying the Characteristics of Human Visual Perception

• Main Authors: TING, CHIH-CHUNG, 丁致中
• Other Authors: CHIU, CHUNG-CHENG
• Format: Others
• Language: zh-TW
• Published: 2016
• Online Access: http://ndltd.ncl.edu.tw/handle/03463415941715340426

博士 === 國防大學理工學院 === 國防科學研究所 === 104 === Image enhancement technologies are primarily to improve the contrast of the images and make them look clearer. In recent years, because the technologies are advancing too fast, image enhancement technologies have been widely used to improve the visual effects of images. Histogram equalization (HE) is one of the most popular image enhancement methods for enhancing image contrast owing to its simplicity and effectiveness. However, HE may cause excessive contrast enhancement and feature loss problems that result in an unnatural look and loss in details of the processed images. As a result, many researchers have proposed various HE-based methods to solve the excessive contrast enhancement problem; however, they have not considered the feature loss problem. This thesis proposes the Visual Contrast Enhancement Algorithm Based on Histogram Equalization (VCEA) and the Improved Visual Contrast Equalization Algorithm Based on Histogram Equalization(IVCEA). VCEA retains the image enhancement effect of HE and solves the excessive contrast enhancement problem by considering human visual perception. Then, it solves the feature loss problem. In addition, it also further enhances the detailed textures of the images. The processed images have better image quality and satisfy human visual perception. According to the characteristics of human visual perception, IVCEA proposes the gap adjustment formula to solve the excessive contrast enhancement problem effectively and to make the enhanced image meet the requirement of human visual perception. Furthermore, IVCEA can also solve the feature loss problem and enhance the detailed textures in the dark regions of the images to improve the quality of the processed images for human visual perception. It is worth noting that these two algorithms solve the feature loss problem which has not been solved for a long time. Experimental results show that the two proposed algorithms outperform HE and other HE-based methods. The most important thing is that the images processed by these two algorithms not only satisfy the human visual perception but also have better visual quality.