一种基于视觉适应性的图像增强算法

模拟真实场景反应在人眼中形成图像的过程,考虑到邻域间干扰、光照不均匀、媒介传递率下降引起的图像退化,提出了一种新的图像增强算法。依据对视觉现象中对比灵敏度、亮度适应能力和侧抑制现象的研究,建立了一种图像退化模型的视觉适应性模型,该模型求解方案包含3个过程:用于消除邻近像素相互影响,给人眼提供更多细节的去卷积过程;据锥细胞亮度适应能力提出的,为其工作提供最佳工作光照条件的亮度调整过程;模拟侧抑制机理提出的对比度拉伸过程。对比现有图像增强算法的增强结果,实验结果表明该方法的视觉适应性解,在增强图像细节、提升对比度、增强结果的视觉愉悦性方面取得了较大进步。

Acquisition of Vision Adaptability Image

Through simulating the process of image construction we propose a new image degradation model in this paper, and mutual effect from adjacent pixels, not well-proportioned illumination, as well as the decline of medium transfer ratio are taken into account in the new model. Based on the research on human visual system, including contrast sensitivity, bright- ness adapting ability and side restraint, a vision adaptability solution for the model is proposed. There three processes are included in the solution: a de-convolution process to eliminate the adjacent pixels'' mutual effect and provide people with more details; an intensity adjustment process to provide the cone cells with a suitable working condition (the intensity adjustment function is proposed according to cone cells brightness adapting ability) and the contrast upgrade process proposed through simulating the side constraint thesis. Compared with some existing image enhancement algorithms, it can be found in the experimental results that the applications of the proposed algorithm yield some promising results in getting more details, upgrading the contrast, and increasing the visual pleasing characters.