Water color exception object extraction method and system based on multi-spectral remote sensing image

Abstract

The invention discloses a water color exception object extraction method and system based on multi-spectral remote sensing. The method comprises the steps of first acquiring the reflectance of a remote sensing image, then carrying out separation of land and water and cloud removing to obtain remote sensing reflectance of a water area portion, and conducting a principal component analysis of the multiband remote sensing reflectance to obtain a first principal component; afterwards, conducting a histogram analysis of a first principal component image, setting a threshold, segmenting the first principal component into a target component and a background component by using the threshold, segmenting the first principal component by using a threshold when the between-class variance of the target and background components is maximum and by setting the multiple, and obtaining a water color exception object portion; and conducting pixel statistics for the water color exception object portion, and obtaining the area and distribution of water color exception objects. The invention can retain the amount of multiband information to the greatest extent, thereby improving the accuracy; and meanwhile, acquires the segmenting threshold completely based on automatic calculation of a computer, thus greatly saving the information extracting time, and achieving simple and accurate extraction of distribution information of the water color exception objects.

Description

technical field [0001] The invention relates to the field of water body monitoring, and more particularly to a method and system for extracting abnormal water color objects based on multi-spectral remote sensing images. Background technique [0002] Objects with abnormal water color, such as the outbreak of Enteromorpha, are a kind of water environment disaster caused by eutrophication in coastal waters. When the abnormal water color objects break out, the transparency of the water body will decrease, and the degradation products after the death of the abnormal water color objects will reduce the dissolved oxygen in the water body, deteriorating the water quality, and seriously affecting the water quality of the water body. , Rapid monitoring, so as to carry out real-time deployment and emergency treatment of water color abnormal object disasters, provide technical support for environmental management and emergency departments, and improve the monitoring ability of sudden wa...

AI Technical Summary

Problems solved by technology

[0004] The thresholds in this method are all set artificially based on experience, and the operation steps are cumbersome, and the selection of thresholds is different due to the number of sample points selected by different personnel, so the area and range of water color anomaly objects are easily caused by human factors The accuracy is low; at the same time, this method only considers the red band and the near-infrared band, and does not consider the spectral information of the water-color anomaly object in other bands, resulting in a large loss of spectral information (and for some water-color anomaly objects, there is no obvious feature bands), which greatly reduces the accuracy of water color anomaly object extraction

In addition, the above method is suitable for the information extraction of water surface conditions completely covered by water color abnormal objects, but for pixels mixed with water bodies and water color abnormal objects (pixels composed of water color abnormal objects), the extraction accuracy is extremely low

Images