Image Processing Method for Intruder Detection around Power Line Towers
Image Processing Method for Intruder Detection around Power Line Towers
复制标题
输电线塔周围入侵者检测的图像处理方法
DOI：
--
复制
发表时间：
1992
期刊：
影响因子：
--
通讯作者：
T. Nagai
中科院分区：
文献类型：
--
作者：
M. Kaneta;H. Kanoh;K. Kanemaru;T. Nagai
提出了一种利用图像处理技术检测电力线塔周围入侵者的方法。室外图像包括导致错误图像处理的各种因素，诸如亮度的快速变化、树叶的卷曲、雾、雨以及由于风引起的电力线导体的错位。这些问题的解决方法如下。首先，使用参考图像和观察图像之间的差的直方图来检测图像的变化，作为入侵者候选。使用外接矩形、数字、面积和重心进一步确定检测到的差异是入侵者。该方法经现场试验验证，入侵检测准确率为82%。
A method of detecting intruders around pow~r line tow~rs using an image processing technique has bcen developed. Outdoor images include a variety of factors leading to erroneous image processing, such as rapid changes in brightness, r~istllng of leaves, mist, rain, and dislocation of power lin~ conductors due to wind. These problems were solved as follows. A change of image is first detected, as an intruder candidate, using a histogram of the differencr between a reference image and an observation image. The detected differences ate further determined to be an intruder using a circumscribed rectangle, a number, an area, and the center of gravity. The method was field tested and the evaluation confirmed that the sr~ccessful intruder detection rate was 82%.