Smoke concentration estimation model training method and device, electronic equipment and medium

Abstract

The invention discloses a smoke concentration estimation model training method and device, electronic equipment and a medium, and belongs to the technical field of fire detection. The method includes: firstly, building a smoke image data set; then executing a first training stage to obtain a first estimation model; then executing a second training stage; randomly selecting a first number of input images, inputting the input images into a first estimation model, randomly extracting a second number of background pixel feature representations from first output features output by the first neural network, and obtaining a background pixel feature set; generating a pseudo tag according to the similarity between the first output feature and the background pixel feature set; calculating a loss function value according to the pseudo label and the output of the first estimation model, and updating parameters of the first estimation model according to the loss function value; and circulating the second training stage to obtain a smoke concentration estimation model meeting the requirements. According to the invention, the smoke concentration estimation model can be efficiently trained, and the performance of the smoke estimation model is improved.

Description

technical field [0001] The invention belongs to the technical field of fire detection, and in particular relates to a training method, device, electronic equipment and medium for a smoke concentration estimation model. Background technique [0002] Smoke and flame detection using video images is a low-cost implementation of flame detection, where smoke is an early manifestation of flames, so monitoring abnormal smoke is an important part of smart cities and firefighting. However, the smoke target is different from common rigid bodies, its edges are blurred and translucent, and it belongs to a special kind of fluid target. Smoke targets are usually mixed with background information, and smoke features are easily affected by ambient light conditions, making it difficult to efficiently detect and label smoke targets. [0003] In the prior art, a neural network model is usually used to estimate the smoke concentration to determine whether a fire occurs. However, the detection ...

AI Technical Summary

Problems solved by technology

[0004] Technical problem: The present invention aims at the problem that it is difficult to effectively train the smoke concentration estimation model in the existing training method of the smoke concentration estimation model, and provides a training method, device, equipment and storage medium of the smoke concentration estimation model to improve the training efficiency. The performance of the smoke density estimation model after

Images