摘要
Harris角点检测算法是图像匹配中使用非常广泛的特征提取算法。针对目前图像处理过程中尺度不变特征点提取的算法实时性较差,算法计算量比较大的问题,在Harris角点检测算法的基础上提出一种简化的算法思想:邻近像素采用对比的方法,从理论上分析算法的性能,实验中保证算法性能的同时实时性得到了提高。
Harris corner detection algorithm is widely used in feature extraction algorithm in image matching. Against the issues of current scale-invariant image processing algorithm for extracting feature points less real-time and the calculation of algorithm larger then ever, proposes an idea of a simplified algorithm based on Harris corner detection algorithm that is take contradi - stinction from adjacent pixels, a theoretical analysis algorithm performance, the experiment at the same time to ensure real-time algorithm performance has been improved.
出处
《现代计算机》
2010年第6期44-46,57,共4页
Modern Computer
基金
上海市教委项目(No.2008097)