摘要
针对数字图像在获取过程中动态范围容易产生线性压缩,导致图像的对比度较低的问题,提出一种照度分割下的多尺度增强算法。根据韦伯定律,将图像分成不同的照度区域分别增强。新算法在区域增强上,提出多尺度下差分图像的自适应权重和来实现,再将不同区域的增强图像线性融合;在尺度的选择上,通过分析尺度在所提方法下对于增强图像的影响特性,对各照度区域选取不同的尺度组合。本文给出了该算法与其他算法的对比效果和评价指标。实验结果表明,该算法在对比度提升的同时,还起到一定的锐化作用,具有良好的增强效果。
In gaining the digital image, its dynamic range often produces linear compression, which leads to low image contrast. To solve this problem, a multi-scale enhancement algorithm based on illuminance partition is proposed. According to the Weber law, the algorithm divides the image into several regions with different illuminances and enhances these regions respectively. The regional enhancement is realized by the adaptive weight sum of differential image with different scales. Then the enhanced images in different regions are linearly fused. For different regions different scale combinations are selected by analyzing the influencing character of different scales on the enhanced image. The proposed algorithm is compared with other algorithms. Results show that this algorithm possesses good enhancement effect, which can improve the image contrast and sharpening.
出处
《吉林大学学报（工学版）》
EI
CAS
CSCD
北大核心
2012年第2期494-498,共5页
Journal of Jilin University:Engineering and Technology Edition
基金
国家自然科学基金项目(50904025)
船舶工业国防科技预研项目(10J3.1.6)
中央高校基本科研业务费专项资金项目(HEUFC100809)
关键词
计算机应用
照度划分
多尺度增强
LIP模型
线性融合
computer application
illuminance partition
multi-scale enhancement
LIP model
linerfusio
