Image Enhancement Based on Histogram Equalizationand Image Fusion

• Main Authors: Chih-Ming Lin, 林志明
• Other Authors: Cheng-Hsiung Hsieh
• Format: Others
• Language: zh-TW
• Published: 2010
• Online Access: http://ndltd.ncl.edu.tw/handle/64228297913611357824

碩士 === 朝陽科技大學 === 資訊工程系碩士班 === 98 === This paper presents an image enhancement approach based on histogram equalization (HE) and image fusion, which is abbreviated as IE/HEIF. The proposed IE/HEIF consists of three stages: First, obtain the complement of original image. Second, fuse the original image and its complement by linear interpolation and apply limited range histogram equalization to the fused image. Third, apply linear interpolation once more to fuse the original image and that obtained in the second stage. To improve the performance of IE/HEIF further, an adaptive scheme based on standard deviation of image blocks is employed for parameter adjustment in the linear interpolation. The approach is called IE/HEIF_std. By this doing, it is expected to have better image contrast and visual quality. To justify the proposed approaches, several examples are provided in the simulation where comparisons with other HE-based approaches are made as well. The results indicate that both IE/HEIF and IE/HEIF_std are able to enhance image contrast and visual quality and better results are generally for IE/HEIF_std as expected. Moreover, the proposed IE/HEIF and IE/HEIF_std have better enhanced images than those HE-based approaches which suffer from over-enhancement and un-natural looking in the resulted images in general.