A Method for Separating Heat Source Regions of Electrical Equipment Based on Improved Seed Filling Algorithm

Abstract

The invention discloses a method for separating heat source regions of electrical equipment based on an improved seed filling algorithm, comprising: (1) using an infrared thermal imaging sensor to monitor the temperature of the electrical equipment, obtaining an infrared thermal image, and further obtaining the grayscale of the infrared thermal imaging (2) Use the bilateral filtering method to separate the grayscale image to obtain high-frequency components and low-frequency components, and then perform image enhancement processing on the high-frequency components and low-frequency components and then fuse them; (3) Use an adaptive method to determine the number of categories The K-means clustering algorithm clusters the grayscale image processed in step (2), separates the background layer of low grayscale, and further highlights the high-frequency subject; (4) uses the improved seed filling algorithm to optimize the seed point, And according to the seed point and the boundary condition of filling, the main hotspot area is divided. The invention can quickly, accurately and reasonably separate the hot spots in the infrared image of the electrical equipment.

Description

technical field [0001] The invention belongs to the field of image segmentation and feature separation and extraction, and in particular relates to a method for separating heat source regions of electrical equipment based on an improved seed filling algorithm. Background technique [0002] The infrared thermal imaging technology has been applied to the fault detection of electrical equipment for a long time, but the traditional thermal fault detection of electrical equipment is done manually, which is not only inefficient but also has great uncertainty in the reliability of the results. In recent years, with the continuous development of artificial intelligence, the fault detection of electrical equipment has gradually shifted from manual with low efficiency and low reliability to efficient and accurate computer. [0003] However, infrared thermal images are different from visible light images. Infrared thermal images have the characteristics of low resolution, low contrast,...

AI Technical Summary

Problems solved by technology

Some basic image segmentation methods, such as the threshold segmentation method, need to face how to calculate and select a reasonable segmentation threshold. The most common method of maximum inter-class variance is also difficult to apply to practice due to the huge amount of calculation. Another example is the edge-based image segmentation method. There is a contradiction between anti-noise performance and detection accuracy when edge detection occurs. If the detection accuracy is too high, false edges will be generated due to noise. If the noise resistance is improved, edge detection omissions and deviations will easily occur.

[0005] "Journal of System Simulation" in June 2017 "Improved 2D OTSU and Adaptive Genetic Algorithm for Infrared Image Segmentation" published a method combining the improved 2D OTSU algorithm and adaptive genetic algorithm, which improved the threshold search speed and infrared The segmentation efficiency of the image, but the genetic algebra at the end of the genetic algorithm is considered to be artificially set according to the complexity of the image. Too much manual optimization is required in terms of accelerated operation and intelligent selection of the genetic algebra, so the segmentation effect is not universal.

Images