CN111076815A

CN111076815A - Hyperspectral image non-uniformity correction method

Info

Publication number: CN111076815A
Application number: CN201911124467.6A
Authority: CN
Inventors: 张长兴; 王跃明; 张东; 何道刚; 王建宇
Original assignee: Shanghai Institute of Technical Physics of CAS
Current assignee: Shanghai Institute of Technical Physics of CAS
Priority date: 2019-11-18
Filing date: 2019-11-18
Publication date: 2020-04-28
Anticipated expiration: 2039-11-18
Also published as: CN111076815B

Abstract

The invention discloses a method for correcting the heterogeneity of a hyperspectral image. The method aims at the difficult problem of image non-uniformity correction caused by lack of non-uniformity correction coefficients or coefficient failure, basically does not cause loss of image information, and has important effects on imager calibration and data preprocessing.

Description

Hyperspectral image non-uniformity correction method

Technical Field

The invention belongs to the field of remote sensing detection and imaging spectrometer data processing, and particularly relates to a method for correcting heterogeneity of a hyperspectral image.

Background

Due to the change of materials, processes and environments of the detector, the response value of the detector is different for uniform radiation or reflection signals, and the image quality and the application effect of the imaging spectrometer are seriously influenced by non-uniformity. The non-uniformity is therefore a necessary element of image processing.

The non-uniformity correction method comprises a statistical matching method based on a laboratory calibration method, image filtering and an image, wherein the laboratory calibration method can achieve high precision, but the problem of data non-uniformity correction is caused by calibration coefficient failure caused by insufficient laboratory calibration and the influence change of a detector due to environment and the like; the image filtering can cause the loss of image information, and the algorithm is complex and the operation speed is slow; the statistical matching method based on the images is mainly based on moment matching, the statistical content is different from the statistical type of the method, most of the statistical content is the mean value, standard deviation, correlation coefficient and the like of the images, and the method has higher requirements on the scene distribution of the images. Therefore, a non-uniformity correction method is needed to solve the problem of non-uniformity correction caused by no scaling coefficient or failure of the scaling coefficient.

Disclosure of Invention

The invention aims to solve the problems that: the method for correcting the image nonuniformity solves the problem of image nonuniformity correction caused by invalid calibration coefficients at present.

The invention comprises the following steps:

(1) acquiring flight image data of a single-imaging spectrometer, performing data preprocessing, and subtracting dark current from all image data to obtain an image data set.

(2) Carrying out distribution statistics on image data acquired by each probe element, wherein the steps are as follows:

(2-1) counting the maximum value, the minimum value and the dynamic range of all image gray values of each probe element;

and (2-2) counting the cumulative histogram of each probe element, and carrying out normalization processing on the cumulative histogram to generate a normalized cumulative histogram.

(3) Generating a radiation correction coefficient, and the steps are as follows;

(3-1) extracting gray values G corresponding to designated normalization values of T groups (T is more than or equal to 2) from the normalized cumulative histogram, and solving the average value H of the gray values G of each row;

(3-2) calculating correction coefficients A and B for the nonuniformity correction using the following equation

Wherein:

is the average of the T sets of gray values G,

the mean value of the mean value H of the T groups, and T is the number of groups extracted in the step (3-1).

(4) And (4) non-uniformity correction, namely correcting all pixels by using the non-uniformity correction coefficient to obtain a non-uniformity correction result.

By the system, the condition that laboratory calibration is lacked and the imager is influenced by environment and the like to change and have calibration coefficient failure can be realized, the problem of image non-uniformity correction can be effectively solved, and compared with the conventional image denoising processing method, the system well keeps the information content of the image.

Drawings

The invention will be further described with reference to the accompanying drawings and examples, in which:

FIG. 1 is a flow chart of a method of non-uniformity correction;

fig. 2 is a comparison graph of the effect of the non-uniformity correction performed by the present invention, wherein (a) is an original image and (b) is a non-uniformity correction result.

Detailed Description

The following describes the non-uniformity correction method in detail with reference to fig. 1 to 2.

(1) Acquiring flight data of a single-time imaging spectrometer, wherein the size of a detector of the imaging spectrometer is 1024 multiplied by 256, the space dimension is 1024, the spectral dimension is 256, and a single flight path is acquiredTaking about 100000 frames of image size, preprocessing data, subtracting dark current from all data to obtain image data set D_n(i,j)，D_n(i, j) represents the image gray values of the ith column, the jth row and the nth frame of the detector, i is more than or equal to 0 and less than 1024, j is more than or equal to 0 and less than 256, and n is more than or equal to 0 and less than 100000; part of the original image is shown in FIG. 2 (a);

(2) carrying out distribution statistics on the gray value of the image collected by each probe element;

(2-1) counting the maximum value Max (i, j), the minimum value Min (i, j) and the dynamic range Rag (i, j) of all collected gray values of each probe element, wherein the Rag (i, j) is Max (i, j) -Min (i, j) + 1;

(2-2) counting the cumulative histogram of each probe element and carrying out normalization processing to obtain a normalized histogram H_r(i,j)，H_rAnd (i, j) represents the ratio of the number of pixels of the ith column and the jth row of the detector, wherein the gray value of the image is less than or equal to r, to the total frame number.

(3) Generating a radiation correction coefficient;

(3-1) extracting 9 groups of characteristic distribution values and solving for H_r(i, j) equals the gray values corresponding to 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, respectively corresponding to G_t(i, j), t ═ 1,2, …, 9; calculating the mean value H of the gray value of each line_t(j) The calculation formula is

(3-3) obtaining correction coefficients A (i, j), B (i, j) for the nonuniformity correction for each probe by the following equation

Wherein:

(4) and (3) non-uniformity correction, namely correcting all pixels by using the non-uniformity correction coefficient to obtain a non-uniformity correction result:

R_n(i,j)＝A(i,j)×D_n(i,j)+B(i,j)

wherein R is_n(i, j) represents the image uniformity correction result of the ith column, jth row and nth frame of the detector, and a partial correction result graph is shown in FIG. 2 (b).

Claims

1. A hyperspectral image nonuniformity correction method is characterized by comprising the following steps:

(1) acquiring flight image data of a single-imaging spectrometer, performing data preprocessing, and subtracting dark current from all image data to obtain an image data set;

(2) the method comprises the following steps of carrying out distribution statistics on image data acquired by each probe element:

(2-2) counting the cumulative histogram of each probe element, and carrying out normalization processing on the cumulative histogram to generate a normalized cumulative histogram;

(3) generating a radiation correction coefficient, which comprises the following specific steps;

(3-1) extracting T groups from the normalized cumulative histogram, wherein T is more than or equal to 2, designating a gray value G corresponding to a normalized value, and solving the average value H of the gray values G of each row;

Wherein:

is the average of the T sets of gray values G,

the mean value of the T groups of mean values H, T is the group number extracted in the step (3-1);