摘要
基于Sobel边缘检测算法处理图像,得到边缘图像以及灰度共生矩阵。将该矩阵的多个统计量记为描述图像的特征向量,并对图像进行检索。实验结果表明,该算法能够大幅缩减检索时间,并且对于纹理较为丰富的图像具有较好的检索效果。
Sobel edge detection is applied to process images to get the image edges and gray level co-occurrence matrix.The statistic parameters of the matrix are taken as the image feature vectors for image retrieval.Experiment results indicate that the algorithm is effective to reduce retrieval time,and for abundant texture image detection.
出处
《长春工业大学学报》
CAS
2017年第2期174-178,共5页
Journal of Changchun University of Technology
基金
青海省创新平台建设专项基金资助项目(2016-ZJ-Y04)
关键词
图像检索
阻塞直方图
SOBEL算子
灰度共生矩阵
image retrieval
blocked histogram
Sobel operator
gray-level co-occurrence matrix
