摘要
针对激光主动成像图像特点,提出一种改进的Canny边缘检测算法。该算法在Canny边缘检测算法的基础上,采用投票中值滤波和提升小波级联的方法代替传统的高斯滤波器;采用3×3邻域的权值自适应梯度计算方法代替原有的2×2邻域差分运算。仿真实验证明:该算法提升了传统Canny算子的抗干扰能力和鲁棒性,能够较好地保存边缘信息,准确检测出图像边缘,并且实时性较好,更具实用性。
According to the image characteristics of laser active imaging,an improved Canny edge detection algorithm was proposed. Based on Canny edge detection,the algorithm used the voting median filtering and lifting wavelet transform to replace the traditional Gaussian filter,and calculated the grades of filtered images by using an adaptive method of 3 × 3 adjacent regions gradient. Simulation results show that： the algorithm can improve the anti-interference capability and robustness of traditional Canny operator,keep well the detail edge information and has fine real-time performance,which makes it more practical.
出处
《电光与控制》
北大核心
2015年第4期27-31,共5页
Electronics Optics & Control
基金
预研基金(9140A010604)
