论文部分内容阅读
采用传统的人工定期巡检的方式进行电力设备状态检测,一旦待检测设备较多,将会耗费大量的人力成本,且无法保证检测的实时性与检测效率,无法满足电力设备安全运行的需要。提出基于嵌入式结合计算机视觉的电气设备状态监控系统设计方法。针对计算机采集到的原始图像进行中值滤波,在最大化降低噪声的同时,更好地保护图像中的电气设备的细节信息,对去噪后的电气设备图像进行边缘增强处理,获取锐化后的电气设备图像,依据马尔可夫随机场纹理分析法原理对电力设备图像进行纹理分析,通过与正常图像模板进行比较得出差异,实现对电力设备当前状态的识别与检测。实验结果表明,利用该方法进行电力设备状态检测,能够对检测结果进行自动识别与分析,极大地提高了检测效率。
Traditional manual inspection of the state of the power equipment state detection, once the equipment to be tested more, it will cost a lot of manpower costs, and can not guarantee the real-time detection and detection efficiency, can not meet the needs of the safe operation of power equipment. A design method of condition monitoring system for electrical equipment based on embedded vision combined with computer vision is proposed. For the original image collected by the computer median filter, to maximize the noise reduction at the same time, to better protect the details of the electrical equipment in the image information to the edge of the electrical equipment image denoising enhanced processing, access to sharpening Based on the principle of Markov random field texture analysis of the power equipment image texture analysis, compared with the normal image template to draw differences to achieve the current status of power equipment identification and detection. The experimental results show that using this method to detect the state of power equipment can automatically identify and analyze the detection results, which greatly improves the detection efficiency.