A shallow sea coral reef remote sensing information automatic extraction method

Abstract

The invention provides a shallow sea coral reef remote sensing information automatic extraction method, and the method comprises the steps: carrying out the preprocessing of a remote sensing image, carrying out the median filtering of the preprocessed image for eliminating salt and pepper noise or spots, smoothing the image to reduce a wrong classification result, and enable the extraction resultto present a better effect. Threshold segmentation is carried out by using a normalized difference water body index (MNDWI), land information is masked, and water body information is extracted. And threshold segmentation is performed by using the green light wave band of the image, open water body information is masked, and reef information is extracted. Threshold segmentation is carried out by utilizing normalized difference vegetation index (NDVI), the NDVI can effectively identify algae information in the coral reef, then the coral reef and the sand beach information are distinguished, thesand beach information is masked, and the coral reef information is extracted. The method provided by the invention is simpler, simplifies the previous relatively complex extraction process, has higher operation efficiency, and has better practicability.

Description

technical field [0001] The invention belongs to the field of automatic extraction of ground objects from remote sensing images, and relates to a method for extracting information on shallow sea coral reefs from remote sensing images, in particular to a method for extracting remote sensing information on shallow sea coral reefs in a layered and stripped manner. Background technique [0002] Coral reefs are an ecosystem of great biodiversity and ecological value. Coral reefs have extremely high primary productivity. The area of ​​coral reefs accounts for about 0.25% of the total ocean area in the world, but they support more than 1 / 4 of marine fish and are important marine fisheries. Because of its own unique growth structure, it has created a colorful coral reef world, which contains rich tourism resources. Coral reefs respond very quickly to seawater temperature, acidity, and seawater pollutants, and can be used as a major biological indicator of the marine environment to r...

AI Technical Summary

Problems solved by technology

This method is labor-intensive and relies on the prior knowledge of the staff

[0007] (3) The supervised classification method is based on the establishment of statistical recognition functions as the theoretical basis and the classification technology based on typical sample training methods. The determination of the classification system and the selection of training samples in this method are all subjective factors. It takes more manpower and time to select and evaluate, and the degree of automation is low

[0009] Generally speaking, the research on remote sensing information extraction methods for coral reefs is relatively scarce, and the research on shallow coral reefs is even more scarce. The extraction accuracy of various methods is different, and its operation is relatively complicated, and the degree of automation is relatively low.

Images