考虑视觉相似性的广角图像背景区分层增强

Layered Background Enhancement of Wide-Angle Image Considering Visual Similarity

为了解决广角图像背景区内容复杂导致的视觉效果差问题,提出一种考虑视觉相似性的广角图像背景区域分层增强方法。首先根据视觉习惯将对图像亮度进行调整,通过相似性像素取值计算出符合人眼需求的相似性曲线,使最初图像转换至模糊域中,并求解得出模糊特征平面,再运用分段函数实现灰度映射增强,同时将增强后的图像空间像素分层映射到对应模糊集合中,运用迭代增强函数求解获取出新模糊集合,并按照用户个人视觉相似性来控制增强迭代计算次数,实现图像背景区增强。仿真结果证明,所提方法可有效提高图像层次感,并表现出较好的对比度和亮度,说明所提方法具有良好的图像增强效果。

The background area of wide-angle image with complex content leads to poor visual effect.Therefore,a hierarchical background enhancement method considering visual similarity is proposed in this paper.Firstly,the brightness of the image was adjusted based on the visual habit.According to the value of similarity pixels,the similarity curve that satisfies the needs of human eyes was calculated,thus the initial image into the fuzzy domain was transformed,and the fuzzy feature plane was obtained.Secondly,the segmentation function was used to enhance the gray mapping.Meanwhile,the enhanced image spatial pixels were mapped to the corresponding fuzzy sets,and then the iterative enhancement function was utilized to obtain the new fuzzy set.The number of enhancement iterations was controlled according to the user’s personal visual similarity.Finally,the background region of the image was enhanced.The simulation results show that the method has excellent contrast and brightness,which implies that it can effectively improve the image hierarchy and enhance the image effect.