人眼视觉拟合函数的小波系数插值图像增强

A wavelet coefficient interpolation image enhancement based on visual fitting function

空域中直方图增强图像容易产生灰度级断层,导致对比度过分增强、图像局部过亮、细节不清晰等问题。提出一种人眼视觉拟合函数的小波系数插值均衡图像增强方法。算法通过二代小波整数变换,对待增强图像作单层小波分解;在小波低频系数中进行均衡、插值、再均衡计算;最后,将新的低频系数重构图像并归一化,通过人眼视觉拟合函数输出图像。实验结果表明,本方法有效解决了增强图像亮度过亮造成细节不清或丢失、灰度层次感不强的问题。本算法增强的图像视觉效果柔和,更适合图像后续处理与分析。

In the airspace,Histogram Enhanced image is prone to cause grayscale faults,which will lead to excessive enhancement of image contrast,partial overlighting and unclear details.A new method of image enhancement based on human visual fitness function is proposed.First of all,the algorithm performs single layer transform on the image through the lifting wavelet.Secondly,it is balanced,interpolated and re-balanced in the low frequency coefficients of wavelet.Finally,the new low frequency coefficients are reconstructed and normalized,and the image is output through the human eye vision fitting function.The experimental results show that this method can effectively solve the problems that the enhanced image brightness is too high,the details are not clear or lost,and the gray level is not strong.The image enhanced by this algorithm has a soft visual effect and is more suitable for image subsequent processing and analysis.