A transmission line component recognition method based on mixed data processing of GPU and CPU

Abstract

The invention discloses a power transmission line part identification method based on GPU and CPU blended data processing. The power transmission line part identification method based on GPU and CPU blended data processing includes the steps: according to the input image folder path, enabling an image in the path to form a data flow file, and according to the size of image data, opening up the memory space for an image processor; extracting the pixel value of the image, and performing memory data exchange between a CPU (Central Processing Unit) and an image processor; according to the size of the original data of the image and the kernel amount, on the basis of guaranteeing that every thread memory is filled, distributing memory for every thread, and guaranteeing the corresponding relationship between the data after processing and the original data at the same time; utilizing a prewitt operator to extract the edge characteristics of the image, and guaranteeing the data consistency at the same time; and according to the obtained edge data through calculation of the image processor, determining the pole or lead part in the image by the CPU. The power transmission line part identification method based on GPU and CPU blended data processing fully utilizes the performance advantages of the CPU and GPU, and improves the execution efficiency of task identification.

Description

technical field [0001] The invention relates to a power transmission line component identification method based on mixed data processing of GPU and CPU. Background technique [0002] With the rapid development of my country's national economy and the increasing scale of urban construction, the increasing demand for power energy is met by adding high-voltage and ultra-high-voltage lines. In order to ensure the stable power supply and safe operation of the entire power grid, routine and special inspections of transmission lines are required. Manual inspections usually rely on inspection personnel to carry relevant equipment (telescopes, infrared imagers, etc.) for observation, which cannot fully observe the operating status of transmission lines and equipment, and is labor-intensive. This traditional inspection mode can no longer meet the inspection needs at this stage. As a new and efficient inspection mode, drones have been introduced into daily inspection work. The UAV in...

AI Technical Summary

Problems solved by technology

The existing processing method uses the CPU to process data alone to complete the component recognition task, but in the current inspection, the number of images and the amount of data obtained in each inspection are large, and it is time-consuming to use only the CPU for data processing. Long, unable to meet the post-data processing requirements of the current inspection

Images