Research On Quality Evaluation Algorithm Of Hyperspectral Remote Sensing Image

With the development of space remote sensing technology,remote sensing technology has become more and more widely used,and has played an increasingly important role in military reconnaissance,resource census,disaster monitoring,environmental monitoring,engineering construction and planning.The quality of remote sensing images directly affects the accuracy of monitoring,prediction and reconnaissance in various fields.Hyperspectral remote sensing images have high spectral resolution,multiple bands and can obtain almost continuous spectral characteristic curves of ground objects.Compared with panchromatic and multi-spectral remote sensing images,hyperspectral remote sensing images can detect more electromagnetic wave response characteristics of the ground object under the same spatial resolution.Previous studies have done a lot of research on the quality evaluation of single-band remote sensing images,but there are few studies on hyperspectral remote sensing images.Therefore,the study on the quality evaluation of hyperspectral remote sensing images has important research and application value.This paper mainly studies the quality evaluation of hyperspectral remote sensing images from the following aspects:(1)Research on noise evaluation algorithm of hyperspectral remote sensing image.The noise evaluation based on local standard deviation,the evaluation of local standard deviation noise based on edge block culling and the noise estimation of decorrelation method are studied in depth.The advantages and disadvantages of each algorithm and the applicable scenarios are compared through experiments.(2)Research on hyperspectral remote sensing image fuzzy degree evaluation algorithm.The ambiguity evaluation based on image histogram and the ambiguity evaluation based on gray gradient co-occurrence matrix are deeply studied.The advantages and disadvantages of each algorithm and the applicable scenarios are compared through experiments.(3)Research on cloud content detection algorithm for hyperspectral remote sensing images.The cloud detection algorithm of remote sensing image based on multi-spectral radiance characteristics is studied in depth.On this basis,a cloud detection algorithm based on dynamic fractal dimension and radiance characteristics is proposed.(4)Research on comprehensive quality evaluation algorithm for hyperspectral remote sensing images.Aiming at the comprehensive evaluation of image quality,the support vector regression method and the integrated decision tree method are used to establish the quality evaluation single model for the training set image with evaluation value.The problem of image quality is easy to over-fitting for single model evaluation.Fusion of hyperspectral image quality evaluation algorithms.(5)Hyperspectral remote sensing image quality evaluation software design.The overall architecture and overall process of hyperspectral remote sensing image quality evaluation software were designed,and the functions and method flow of noise evaluation,fuzzy degree evaluation,cloud content detection and comprehensive evaluation module were introduced in detail.