蜣螂优化算法在 Canny 边缘检测算法中的应用

针对传统 Canny 边缘检测需要手动选取阈值以及不能有效提取边缘轮廓的问题，提出了一种基于改进的蜣螂优化算 法(DBO) 来优化 Canny 算子的边缘检测算法。首先通过快速引导滤波代替传统高斯滤波对图像进行保边去噪；其次用4方 向的 Sobel 模板来计算图像的梯度幅值和梯度方向；最后利用蜣螂优化算法优化的二维大津法自适应获取高低阈值。针对蜣 螂优化算法种群多样性不强问题，提出用 tent 映射初始化种群；为了提高算法跳出局部最优的能力，采用精英差分变异策略 对最优蜣螂个体进行变异扰动。实验结果表明，在边缘准确度和连接性上，该算法与传统 Canny 边缘检测算法对比有一定程 度的提升，能够有效提取图像的边缘轮廓，提高了Canny 边缘检测的边缘连接性，具有一定的实用性。

Application of dung beetle optimization algorithm in Canny edge detection algorithm

To solve the problem that traditional Canny edge detection requires manual threshold selection and can not effectively extract edge contour,an improved dung beetle optimization algorithm(DBO)is proposed to optimize the edge detection algorithm of Canny operator.Firstly,the image is denoised by fast guided filtering instead of traditional Gaussian filtering.Secondly,a 4-direction Sobel template is used to calculate the gradient amplitude and gradient direction of the image.Finally,the high and low thresholds are obtained adaptively by using the two-dimensional Otsu method optimized by dung beetle optimization algorithm.Aiming at the problem that the population diversity of the dung beetle optimization algorithm is not strong,this paper proposes to initialize the population by tent mapping.In order to improve the ability of the algorithm to jump out of the local optimum,the elite differential variation strategy is used to carry out variation disturbance on the optimal dung beetle.The experimental results show that in terms of edge accuracy and connectivity,the algorithm has a certain degree of improvement compared with the traditional Canny edge detection algorithm,which can effectively extract the edge contour of the image and improve the edge connectivity of Canny edge detection,which has certain practicality.