Power transformation field operation illegal behavior detection method

Abstract

The invention provides a power transformation field operation illegal behavior detection method, which comprises the following steps of: firstly, acquiring a power transformation field operation illegal behavior image, manually segmenting an operation related target, calibrating an illegal behavior category to which the image belongs, and establishing a power transformation field target segmentation data set and an illegal behavior classification data set; then, training and classifying a U-Net model based on the target segmentation data set, and acquiring a segmentation result of a power transformation site target; in order to reduce the interference of an image background area and an irrelevant object, performing AND operation on the segmented binary target image and an original image to obtain a grayscale image of the target area; and finally, performing classification by using a CNN model according to the grayscale image of the target area, and acquiring a final illegal behavior detection result. By applying the power transformation field operation illegal behavior detection method provided by the invention, the interference of an image background area and an unrelated object can be reduced, the power transformation field operation illegal behavior can be automatically and effectively detected, and the risk prevention and control capability is improved.

Description

technical field [0001] The invention relates to the field of digital image processing, and relates to artificial intelligence and deep learning. Background technique [0002] With the rapid development of the State Grid and the rapid expansion of the coverage of substation equipment, the labor intensity of the staff in the substation has increased significantly. However, due to the lack of safety awareness of the staff and the impact of severe weather conditions, frequent violations of regulations have occurred, and the risk of substation operations has shown a rising trend. For example, forgetting to wear safety helmets, not using safety belts when climbing high, smoking in the work area, operating outdoor high-voltage equipment without wearing insulating boots, etc. In particular, the installation, commissioning, and maintenance of some electrical equipment require staff to work at high altitudes, and there are great potential safety hazards in illegal operations. Theref...

AI Technical Summary

Problems solved by technology

This method is time-consuming and labor-intensive, and the long-term inspection and monitoring can easily make people tired and boring. However, the existing deep learning-based violation detection methods are generally based on pre-processed images, such as grayscale transformation, rotation, and adding noise. Perform training and testing with little regard for the impact of image background or irrelevant objects on detection performance

Images