Automatic suburb village ternary structure extraction method based on DMSP night light image

Abstract

The invention provides an automatic suburb village ternary structure extraction method based on a DMSP night light image. The method comprises the following steps: acquiring first noctilucent lamp data, water body data and fire point data; performing grid multiplication and mask erasing on the first noctilucent lamp data to obtain second noctilucent lamp data; cutting the second noctilucent lamp data to obtain first target area noctilucent lamp data; obtaining a quantile curve according to the first target area noctilucent lamp data, and respectively calculating first, second and third gradingthresholds according to the maximum value of the distance from the quantile curve to the reference line; and respectively carrying out grid re-classification on the first, second and third target areas by utilizing the first, second and third grading thresholds to classify the first, second and third target areas into cities, suburbs, villages and others. According to the method, automatic quantile curve analysis is carried out on the basis of the DMSP night light image, extraction of a suburb village ternary internal fine structure can be achieved, the threshold value is automatically selected from DMSP night light brightness distribution, and errors caused by introduction of auxiliary data and subjective factors are avoided.

Description

technical field [0001] The present invention relates to the technical field of image and video recognition, and more specifically, to a method for automatically extracting the ternary structure of suburban villages based on DMSP night light images. Background technique [0002] The application of satellite remote sensing data of night light is more and more extensive, and it has become an important data source for urbanization and land use monitoring research. DMSP night light data (Defense Meteorological Satellite Program, DMSP) is one of the most widely used night light data, which uses the equipped linear scanning business system to form grayscale images through visible light-near infrared bands. The spectral resolution is 8 bits, and the gray value range is 0-63. Compared with other night light satellites, the amount of data is small, the image processing is relatively simple, and it has obvious advantages in long-term sequence; in addition, the night light data is more...

AI Technical Summary

Problems solved by technology

[0003]However, the existing night light impact extraction technology solutions have the following defects: rely on empirical knowledge and auxiliary data to determine the gray threshold of night light images, and then determine the urban area method, which introduces human subjective factors and errors in auxiliary data, and the accuracy is insufficient; the auxiliary data and empirical knowledge introduced are not universal, and the corresponding empirical knowledge and required auxiliary data are different for different regions. Therefore, this method of adding experi

Images