Smoke and fire point positioning method based on video monitoring

Abstract

The invention discloses a smoke and fire point positioning method based on video monitoring, and the method comprises the following steps: taking a frame of image from video monitoring equipment at equal intervals per second, generating a continuous image sequence, carrying out the regular partitioning of the image sequence according to time, forming image blocks, and calculating the signal-to-noise ratio of each image; setting a threshold value by adopting a background difference method according to the change sequence of the signal-to-noise ratio; according to the signal-to-noise ratio of the image and the threshold value, finding out a changed image block, and determining the changed image block as a suspected smoke image; extracting an LBP feature vector of the suspected smoke image, inputting the LBP feature vector into a trained SVM classification model, classifying the suspected smoke image, and distinguishing a smoke image from a non-smoke image; and calculating the position of the smoke image by using a single-point positioning method, and determining the coordinate of a smoke point by combining forestry compartment data. According to the method disclosed by the invention, video monitoring is utilized to identify the smoke area, the secret-related problem is avoided, and deep learning is utilized to identify smoke, so that the identification accuracy can be improved.

Description

technical field [0001] The invention relates to a method for locating pyrotechnic points based on video surveillance, in particular. Background technique [0002] The resources in the forest are one of our important natural resources and play an important role in the development of human life and society. It can maintain biodiversity, maintain soil and water stability, and maintain the carbon balance in nature. Therefore, protecting forests from fires has significant economic and social benefits. At present, fire monitoring has been greatly improved by using the combination of civil air defense and technical defense in ground monitoring, but how to accurately determine the location of the fire in the later stage of the fire has become one of the technologies that need to be researched next. At present, manual observation and positioning, two-point intersection positioning, laser ranging positioning, high-precision DEM data perspective analysis and positioning are mainly use...

AI Technical Summary

Problems solved by technology

[0003] The artificial lookout observation method needs to be familiar with the terrain of the monitoring area. During the forest fire prevention period, it is necessary to bring wireless communication equipment, handheld GPS, compass and other equipment to observe the fire situation and determine the location of the fire situation. It can determine the horizontal position but cannot accurately determine the distance and depth.

The positioning accuracy of two-point crossing can be guaranteed, but it cannot be realized at the edge of the monitoring area, and at least two monitoring devices need overlapping coverage to achieve it. Such an improvement in positioning accuracy is based on reducing the coverage of the equipment, resulting in a large investment

Laser ranging and positioning will increase the cost of equipment and the effective distance is short, which cannot meet the needs of long-distance positioning

Using high-precision DEM data can realize high-precision spatial perspective analysis and calculate the location of the fire point, but it cannot effectively avoid issues such as confidentiality

Images