一种基于人眼视觉特性的Curvelet域紫外图像增强算法

An ultraviolet image enhancement algorithm in Curvelet domain based on human visual characteristics

针对紫外图像本身存在的对比度低、成像不清晰、整体成像过暗或过亮且受噪声干扰等特点,从人眼视觉的掩蔽特性出发,提出了一种基于人眼视觉特性的Curvelet域紫外图像增强算法。通过Curvelet变换将紫外图像信息分解到不同尺度、不同方向子带,再依据各子带建立HVS模型,利用其遮蔽性对高频分量进行非线性增益来提高图像对比度,同时非线性灰度拉伸低频分量可以提高图像整体亮度;最终将处理过的高低频分量通过Curvelet逆向重建,得到增强后的紫外图像。实验结果表明:增强后的图像符合人眼视觉对图像信息的提取,突出了边缘和目标的轮廓信息,在主观视觉和定量指标方面都具有明显的效果。

Aiming at the characteristics such as low contrast,unclear imaging,over-dark or over-bright overall imaging and noise interference in ultraviolet image itself,a Curvelet domain ultraviolet image enhancement algorithm was proposed based on the masking characteristics of human vision. Firstly,the ultraviolet image information was decomposed to different scales and direction sub-bands by Curvelet transform. Secondly,the HVS model was established according to each sub-band,and with its masking characteristics,the nonlinear gain of the high frequency component was used to improve the image contrast. Meanwhile,the low frequency component of nonlinear grayscale stretching could improve the overall brightness of the image. Finally,the processed high and low frequency components were reconstructed by Curvelet backward to obtain enhanced ultraviolet images. The experimental results show that the enhanced image accords with the extraction of image information from human vision and highlights the contour information of edge and target. In addition,it has obvious effect in subjective vision and quantitative index.