Tidal bore monitoring method based on machine vision

Abstract

The invention discloses a tidal bore monitoring method based on machine vision. The tidal bore monitoring method comprises the following steps: acquiring a water surface video sequence image by utilizing monitoring equipment; performing background elimination processing on the acquired water surface video sequence image to obtain a foreground image; performing foreground ripple noise suppression processing on the foreground image; carrying out tide head line detection on the foreground image subjected to foreground ripple noise suppression processing; and based on the acquired tide head line information, performing measurement and speed estimation on the distance of the tide head line. The method can be used for online monitoring and early warning of the tidal bore, has the advantages of being high in program efficiency, obvious in background interference suppression, prominent in tide head line outline, high in detection precision, high in real-time performance and the like, solves aseries of problems existing in the prior art, and has a better tidal bore monitoring effect.

Description

technical field [0001] The invention belongs to the field of intelligent water conservancy, and relates to a tidal surge monitoring method, in particular to a machine vision-based tidal surge monitoring method. Background technique [0002] After the tidal wave enters the estuary, the deformation is further intensified, and under certain conditions, the peak line of the high tide wave front with a sudden rise in the water level is formed, which is the tidal surge. Tidal surge is a wave train that advances to the upper reaches of the estuary. When the intensity is relatively small, the surface is not broken, which is a wavy tide. When the strong tide meets the ebb tide, the tide breaks and a large amount of air is drawn in, forming a milky white water-air mixture at the head of the tide. However, while it is spectacular, there are also some safety hazards in the tidal surge, which brings a lot of inconvenience to the production and tide watchers along the coast. Taking my c...

AI Technical Summary

Problems solved by technology

[0003] The traditional tide monitoring methods mainly include contact type and non-contact type. The contact type measurement method uses water level gauge and flow velocity flow meter, and uses the rising water level when the tide arrives, the direction of the tide is opposite to that of the river, and the speed of the tide, etc. However, the disadvantage of this kind of contact equipment is that a large amount of sediment will be brought in by the tidal surge, resulting in large measurement errors. At the same time, the long-term contact between the equipment and the tidal water will be continuously corroded and cause system failure.

Ji Sufang and others proposed a method of detecting the tide line using a single frame picture. This method is simple and fast, and can extract the tide line in a picture with a simple background. However, for a complex background environment, this method cannot obtain accurate tide lines. The head line; Zeng Xu et al. proposed a tide detection method based on video recognition. The background modeling of the water surface is carried out through the mixed Gaussian model, and then the tide head line is extracted by the difference operation between the frame to be tested and the background image. This method is suitable for static background It has a good detection effect, but for scenes with large water surface ripples and complex backgrounds, there will be more noise formed by other moving objects, resulting in inaccurate tide line extraction; Gao Peng et al. proposed to use background modeling and particle filtering The model detects and tracks tide surges. This algorithm can realize tide surge detection and dynamic tracking, but it is highly dependent on parameters and does not have the ability of real-time detection, etc.

[0007] Since the viewing angle of the surveillance camera includes not only land, but also complex water surfaces, and tides are non-rigid targets, these factors make video-based tide monitoring research very difficult

Although in the field of moving target detection research, experts and scholars have proposed many algorithms, but they are only applicable to specific environments, so further research is needed for tide surge detection

Images