基于图像识别的保护压板投退状态辨识方法

State Identification of Relaying Plate Based on Image Recognition

通过图像识别方法获取电力继电保护压板投退状态是压板投退管理的有效措施。提出了保护压板投退状态辨识系统,其硬件由计算机、摄像机、显示屏集成嵌入终端构成。嵌入终端通过图像采集、图像去噪处理、目标边缘检测、布局排列分析、状态特征判别等关键步骤完成对压板投退状态的智能识别。采用3/4去噪聚类法,能够有效消除椒盐噪声并实现聚类。系统中设计的基于"块矩阵"的排列分析算法能快速准确地实现目标图像的分割。算例分析结果表明了本文提出算法的有效性和可行性。

Image recognition method, which is used to obtain the on/off states of power relay protection pressed board, is an effective measures for the operating management of press board. The paper proposes the identification system for the identification of the on/off states of power relay protection pressed board. Its hardware is composed of computer, camera, display screen, which is embedded in the terminal. Intelligent recognition of the on/off states of the pressed board of the embedded terminal is accomplished by completing these key steps, such as image acquisition, image denoising processing, target edge detection, layout analysis, state feature discrimination, etc. The paper uses 3/4 denoising clustering method, which can remove the salt and pepper noise effectively and achieve clustering. The permutation analysis algorithm in the system design, which is based on the block matrix, can achieve the segmentation of target image quickly and accurately. The effectiveness and feasibility of the algorithm is proved by the results of the numerical examples.