A large number of aerial images are generated in the inspection of transmission lines. Due to the different sizes of conductors, insulators, shock absorbers and other components, the problem of rejection is easy to occur in the detection of external damage hazards, which affects the overall identification effect of lines. In this paper, a method to identify the hidden danger of external damage of high voltage overhead transmission lines is designed. From the three aspects of color, texture and shape, multi feature extraction is carried out on the line image to represent the location information of external damage hazards. The line image to be recognized is divided into uniform grids, and the coordinates and confidence of sub image blocks are used. Calculate the prediction accuracy of each region, filter the sub region with the largest value as the candidate target region, and output the corresponding prediction box. The deep transfer learning model is used to identify the hidden danger of external damage of the line, and the difference measurement index is used to evaluate the input feature vector. The difference feature results are fed back to the convolution layer to strengthen the interpretation ability of the difference information. The test results show that the design method improves the ability of image feature extraction, has high recognition accuracy, and provides decision-making basis for intelligent line early warning.

The correlation filtering algorithm of infrared spectral data for dim and small target tracking is studied to improve the tracking accuracy of small and weak targets and to track small and weak targets in real time. After the image noise reduction processing by the mean shift filtering algorithm, the infrared small and weak target image data model is constructed by using the denoised infrared small and weak target image. And the brightness value and position of unknown small and weak targets are obtained. The tracking and measurement model of small and weak targets is built. The joint probabilistic data association algorithm is used to calculate the probability that each measurement is associated with its possible source targets, and the particle filter is used to update the tracking status of small and weak targets to achieve real-time tracking of small and weak targets. The experimental results show that the algorithm can enhance the edge contour information of small and weak images, so as to accurately track small and weak targets moving in any track, and has good real-time tracking and accuracy. There is a small deviation between the tracking track of weak and small targets tracked by the algorithm and the actual track, and the root mean square difference of tracking weak and small targets is within 2 pixels. In addition, the detection probability of detecting weak and small targets is less affected by SNR environmental factors.

The existing transmission line fault early warning can’t automatically identify the abnormal area, resulting in poor early warning effect for the line crossing and line touching fault. Based on this, an early-warning method of power transmission line crossing and collision fault based on multi-source data is proposed. Combined with the feature extraction method of multi-source data fusion, based on the analysis of the characteristics of the stock index of power grid transmission line crossing and touching, the leakage current acquisition model is constructed. Based on the model, the abnormal data of power grid are collected, and the characteristics of power grid transmission line crossing and touching are mined and identified. Then the texture features are obtained by using the local binary pattern operator to simplify the early warning steps of power grid transmission line crossing and touching. The experimental results show that the multi-source data-based power grid transmission line crossing and collision fault early warning is more practical than the traditional methods. In the process of practical application, the effect of abnormal data prediction and collection is significantly better. It can carry out fault early warning more quickly and accurately, and fully meet the research requirements.

No abstract