| 基于CNN的彩色图像边缘提取方法 |
| |
| 引用本文： | 薛菲菲,胡燕翔,徐江涛. 基于CNN的彩色图像边缘提取方法[J]. 天津师范大学学报(自然科学版), 2013, 0(2): 47-50 |
| |
| 作者姓名： | 薛菲菲 胡燕翔 徐江涛 |
| |
| 作者单位： | [1]天津师范大学计算机与信息工程学院,天津300387 [2]天津大学电子与信息工程学院,天津300072 |
| |
| 基金项目： | 国家自然科学基金资助项目（61274021） |
| |
| 摘 要： | 根据人眼对R、G和B的敏感特性,提出一种基于视觉感知的矢量角度彩色空间距离度量方法,较传统方法补偿了RGB颜色空间的不均匀性,有效提高了RGB颜色空间内色差度量的精确程度.对基于细胞神经网络(Cellular Neural Network,CNN)进行的彩色图像边缘提取进行理论分析,确定与其功能相适应的最佳模板参数.实验结果表明:该算法不仅实现了彩色图像的边缘提取,同时提取效果更全面,更符合人眼的视觉效果,具有较强的鲁棒性.
|
| 关 键 词： | 边缘提取 细胞神经网络(CNN) RGB颜色空间 颜色空间距离 鲁棒性 |
Edge detection of color image based on CNN |
| |
| XUE Feifei,HU Yanxiang,XU Jiangtao. Edge detection of color image based on CNN[J]. Journal of Tianjin Normal University(Natural Science Edition), 2013, 0(2): 47-50 |
| |
| Authors: | XUE Feifei HU Yanxiang XU Jiangtao |
| |
| Affiliation: | 1. College of Computer and Engineering Information, 2. College of Electronic and Engineering Information, Tianjin Normal University, Tianjin 300387, China Tianjin University, Tianjin 300072, China) |
| |
| Abstract: | According to the sensitivity of human eyes to R, G and B, a distance measurement method based on angle of vectors for color space is proposed based on visual perception. This method improves the ununiformity of RGB color space compared with the traditional methods, and effectively increases the precision of color difference measurement in RGB color space. Edge detection of color image based on cellular neural network （CNN） is analyzed theoretically, and the opti- mal template parameters fit for its function are determined. The experimental results indicate that this method accomplishes the edge detection of color image, obtains the more comprehensive effect of detection, is more in line with the perception of the human eyes, and has extensive robustness. |
| |
| Keywords: | edge detection CNN RGB color space color space dlstance robustness |
| 本文献已被 CNKI 维普 等数据库收录！ |
|
