A neural network smoke image classification method fusing dark channels

Abstract

The invention provides a neural network smoke image classification method fusing dark channels. The method comprises the following steps: preparing two types of samples, namely a smoke image and a non-smoke image, normalizing the samples into the same size, carrying out dark channel processing on all sample images, and dividing an original image and a corresponding dark channel image into a training set, a verification set and a test set to serve as data input of subsequent network training; secondly, training the data set by using the designed dual-channel convolutional neural network, addinga residual block to the first channel network to improve the classification performance, and inputting an RGB original image to extract a generalization feature in the original image; the second channel adopts an improved AlexNet simplified network and inputs a dark channel image to extract detail features of smoke in the dark channel; the two channels are trained respectively, and finally carrying out feature fusion to generate a training model to classify the images. The result shows that the method effectively improves the accuracy of smoke image classification.

Description

technical field [0001] The invention relates to a neural network smog image classification method fused with dark channels, which belongs to the fields of image processing, machine vision, deep learning and environmental protection. Background technique [0002] Fire accidents will inevitably pose a threat to people's lives and property. Traditional fire detection generally uses accurate sensors based on particle sampling, temperature sampling, relative humidity sampling and smoke analysis. Although these sensors are low in cost and simple in principle, because these sensors need to be in physical contact with the gas around the combustion products, they cannot detect fires outside the surroundings of the sensors, so that the occurrence of fires cannot be avoided in time. Studies have found that smoke often appears earlier than open flames, and traditional smoke detectors have similar principles to fire detectors, with longer delays and lower efficiency in larger space envi...

AI Technical Summary

Problems solved by technology

[0004] Existing feature extraction methods have not achieved satisfactory results in smoke detection, mainly because smoke varies greatly in color and shape

In addition, smoke blurs the visual scene, causing the extracted features to be unstable

Therefore, accurate detection of smoke from images remains a challenging task

Images