Multispectral imaging smoke detection method and system

Abstract

The invention discloses a multispectral imaging smoke detection method and system. The detection system comprises a multispectral data acquisition module, a single-chip microcomputer image processingmodule, a mobile network transmission module and a man-machine interaction module. The multispectral data acquisition module is used for acquiring a visible light imaging image and an infrared spectrum imaging image, the single-chip microcomputer image processing module is used for processing acquired image data, the mobile network transmission module is used for sending a data processing result and performing information intercommunication with a single-chip microcomputer, and the man-machine interaction module is used for interaction between a user and a system. The detection method comprises a visible light image processing method which comprises the steps of image feature extraction based on a deep neural network, self-adaptive generation of a target detection area through a segmentation method, and smoke positioning and discrimination. The infrared spectrum image processing method comprises infrared image preprocessing and an ignition point detection algorithm based on a frame difference method. According to the invention, the accuracy of smoke detection is improved, and the method has practical significance for the intelligentization of environmental protection monitoring.

Description

technical field [0001] The invention belongs to the field of environmental protection, and more specifically relates to a multi-spectral imaging smoke detection method and system. Background technique [0002] Smog is an important monitoring indicator in the field of environmental protection. The monitoring and evaluation of smog will be involved in determining the air quality or the degree of migration in a certain place. There are several traditional smoke detection methods, such as manual inspection, drone monitoring, etc. Traditionally, manual on-the-spot inspections are time-consuming and labor-intensive, and cannot provide timely feedback. The human cost of monitoring is very high; although UAV monitoring can reduce part of the labor cost, the hardware cost increases, and it cannot be controlled over a long distance. Manual identification is still time-consuming and labor-intensive. In response to the national call to build an environmentally friendly society, smoke ...

AI Technical Summary

Problems solved by technology

In this method, the number of layers of the neural network used for recognition is relatively shallow, and the expressiveness of the target with complex features such as smoke is not strong enough. Therefore, it is necessary to preprocess the image before inputting it into the neural network to obtain better results.

In this way, the calculation load of the preprocessed image is large, and the recognition speed is slow.

Images