中文摘要：

为了使彩色图像的边缘检测器更符合人眼对图像信息的分辨情况,防止视觉不敏感区域的边缘的过检测问题,该文提出一种自适应色差阈值的估计方法并与不同的色彩梯度算子结合应用于彩色图像的边缘检测中。构建包括亮度掩模与对比灵敏度的局部色差可视阈值的权重因子,结合局部背景亮度以及亮度与色彩的空间频率对人眼视觉的影响。利用信噪比（SNR）,Pratt因子与时间复杂度对提出的算法的抗噪性与边缘定位的准确性以及时间代价进行定量评价,表明该算法能准确检测出图像边缘且有效地抵抗噪声对图像的干扰。

英文摘要：

In order to make the color edge detector considering the discrimination of human eyes, avoiding the perceptually insigni cant edges being over-detected, a novel color gradient based color-edge detection algorithm based on the adaptive perceptual color difference is proposed. A weighting factor including luminance masking effect and contrast sensitivity function is proposed, which combine the influence of local changes in luminance and spatial frequency to human visual system. The SNR and the quality factor Pratt of edge detection evaluation and the time complexity show that the proposed method can detect the color image edge accurately and against the effect of noise.