Human body target feature detection algorithm for power substations by using infrared thermal imaging and visible light imaging technology fusion

Abstract

The invention provides a human body target feature detection algorithm for power substations by using infrared thermal imaging and visible light imaging technology fusion. The human body target feature detection algorithm includes a step S1 of acquiring a visible light image captured by a visible light camera and an infrared thermal imaging image captured by an infrared camera and pre-processing the images respectively; a step S2 of detecting a pedestrian for the visible light image to obtain a visible light image pedestrian detection region A; a step S3 of detecting a pedestrian for the infrared thermal imaging image to obtain an infrared thermal imaging image pedestrian detection region B; and a step S4 of comparing the visible light image pedestrian detection region A with the infrared thermal imaging image pedestrian detection region B, and if the overlap area exceeds a set overlap area threshold, then indicating that a pedestrian target is detected. According to the invention, external influence factors can be well excluded and the accuracy and applicability of the detection can be improved to the maximum extent.

Description

technical field [0001] The invention relates to a human body target detection method, in particular to a human body target detection method combining infrared thermal imaging and visible light imaging technologies. Background technique [0002] The video technology of the substation provides a technical means for the pedestrian visualization of the remote monitoring substation, and provides a visual monitoring and auxiliary decision-making basis for the dispatching operation of the substation. With the continuous development of networked, high-definition, and intelligent technology of video surveillance, the application of video surveillance technology in power grid security and production operation is deepening and extensive. The video surveillance technology used in traditional methods provides remote The visual monitoring method provides a reliable and effective confirmation method for remotely confirming whether there are pedestrians, staff and the status of the staff in...

AI Technical Summary

Problems solved by technology

[0003] There are some shortcomings in the existing pedestrian detection methods in substations. If the commonly used template matching method is used for detection, it is easy to be affected by external light and the image of the camera that captures the image. At the same time, due to the large number of pedestrian-like devices in the substation, the appearance of the equipment and shapes are inconsistent, and different models of equipment from different manufacturers also have certain differences, which can easily interfere with pedestrian detection. Therefore, when the template matching method is used, there will be strict restrictions on the monitoring scene. There are many shortcomings in generalizability and other aspects

Images