摘要
针对热红外图像目标与背景区分不明显、效果模糊,以及传统的Roberts、Sobel、Canny等边缘检测方法难以取得理想检测效果的特点,以玉米植株为测试对象,首次将蚁群优化算法应用于热红外图像边缘检测。该算法由初始化过程开始,进行N步迭代构造信息素矩阵,然后执行信息素过更新过程,最后图像边缘由决策过程给出。仿真实验结果表明,该算法与传统边缘检测算法相比,能够较好地得到边缘检测结果,可为农作物热红外图像处理提供一种新的方法。
The thermal infrared image often has the drawbacks of inconspicuous distinction between target and background, fuzzy effect. The traditional edge detection methods, such as Robert method, Sobel method, Canny method, are difficult to obtain satisfactory results. In order to solve the problems, the corn plant as the test object, the ant colony optimization algorithm is applied for the first time thermal infrared image edge detection. The algorithm begins by initializing process, carried out N-step iterative constructing pheromone matrix, and then performing the pheromone update process, finally, image edge given by the decision making process. The simulation and experimental results show that this method can accurately detect the target edge and it is better than the traditional edge detection. Provided a new method for crop thermal infrared image processing.
出处
《农机化研究》
北大核心
2015年第6期49-52,共4页
Journal of Agricultural Mechanization Research
基金
北京市自然科学基金项目(4142019)
北京市科技新星计划项目(Z111105054511051)
关键词
蚁群优化算法
玉米植株
热红外图像
边缘检测
ant colony optimization
corn plant
thermal infrared image
edge detection