摘要
提出一种自然场景下基于超复数频域Steerable滤波器的电力输电线检测算法.该算法首先利用二维高斯函数构造超复数Steerable正交对滤波器,然后将构造的滤波器与彩色电力输电线图像进行不同方向的脊能量检测,进而获得输电线脊能量最大方向响应,最后得到电力输电线检测图.实验结果表明,该算法具有较好的方向可控性,对输电线的检测比其他算法精度高且速度快,具有一定的理论和实际应用价值.
A detection algorithm of electric power transmission lines in natural environment is proposed based on hypercomplex frequency-domain steerable filter.In this paper,a hypercomplex quadrature pair of steerable filter is constructed with two-dimensional Gaussian function first.Then the ridge energy detection of the color image of the transmission lines and the constructed filters is conducted in various direction,so that the maximum directional response of the ridge energy of the transmission lines is obtained.Finally,the detection graph of the electric power transmission lines obtained.It is shown by experimental result that this algorithm has a better directional steerability and the detection of transmission lines with this algorithm is more accurate and faster than other algorithms,having certain theoretical and practical application value.
出处
《兰州理工大学学报》
CAS
北大核心
2014年第5期82-85,共4页
Journal of Lanzhou University of Technology
基金
国家自然科学基金(61365003)
国家自然科学基金青年项目(61302116)
甘肃省自然科学基金(1308RJZA27)
甘肃省财政厅项目(2013.116)
