基于融合SIFT算法的光伏电站安监图像精准识别技术研究

Research on high-precision recognition technology of photovoltaic power station safety supervision image based on SIFT algorithm

随着我国“双碳”工作的不断推进,光伏电站的规模逐渐扩大。为解决光伏电站图像安防技术中存在的拍摄角度、光线影响大,识别准确率低的问题,提出了一种基于SIFT的光伏电站图像安防技术。通过SIFT算法提取光伏电站图像特征点,解决了光线、角度带来的识别误差问题;通过卷积神经网络方法准确的识别了光伏电站车辆、人员、动物、积雪等安防问题。该方法在某光伏电站进行应用,其光伏电站图像安防识别准确率平均为99.7%。所提方法能有效提高光伏电站图像安防识别准确率。

With the continuous advancement of China’s“dual carbon”work,the scale of photovoltaic power stations has gradually expanded.In order to solve the problems of large influence of shooting angle and light and low recognition accuracy in the image security technology of photovoltaic power station,an image security technology of photovoltaic power station based on SIFT was proposed.The SIFT algorithm was used to extract the image feature points of the photovoltaic power station,which solved the problem of recognition errors caused by light and angle,and the convolutional neural network method was used to accurately identify the security problems of the photovoltaic power station,such as vehicles,personnel,animals,and snow.The method was applied in a photovoltaic power station,and the average accuracy of image security recognition of photovoltaic power station was 99.7%.The proposed method could effectively improve the accuracy of image security recognition of photovoltaic power station.