摘要
提出一种基于可变性图谱局部约束的图像特征提取算法。运用3次样条插值算法逐行扫描特征模糊区域的突变像素点并进行替换,对每一个像素的灰度值与该行的灰度中值进行比较,对像素的分割进行有效的约束,保证模糊的突变像素点被分割算法有效排除。仿真实验结果表明,提出的改进算法能够有效提取出图像中的有效信息,同时节约了算法的执行时间,降低了复杂度。
Put forward a kind of based on local constraints variability map image feature extraction algorithm is proposed. Using three spline interpolation algorithm implementing mutation characteristics of fuzzy area pixe! and replaced. Of each pixel gray value and its value in the comparison of the segmentation of pixel effective constraint. Ensure mutations fuzzy pixel is segmentation algorithm effectively ruled out. The simulation experimental resuhs show that the improved algo- rithm can effectively extract the image of effective information, at the same time to save the execution time of the algorithm, and reduce the complexity.
出处
《科技通报》
北大核心
2013年第12期160-162,共3页
Bulletin of Science and Technology
基金
河南省科技厅科技攻关项目(112102210233)
关键词
特征提取
图像突变像素
约束
feature extraction
mutations pixel image
constraints
