CN109657637B - Method for distinguishing hematite in different rocks by using CASI image - Google Patents

Abstract

Translated fromChinese

本发明属于高光谱遥感矿物填图技术领域，具体涉及一种利用CASI图像区别不同岩石中赤铁矿的方法，步骤一：预处理CASI图像，对预处理CASI图像进行包络线去除，获得包络线去除后的CASI图像；步骤二：对步骤一中所述的包络线去除后的CASI图像进行取子集操作，获得光谱波段在540nm处的图像子集；步骤三：对步骤二中所述的540nm处的图像子集进行中值滤波和像元分类；步骤四：分析判断像元分类结果。The invention belongs to the technical field of hyperspectral remote sensing mineral mapping, and in particular relates to a method for distinguishing hematite in different rocks by using CASI images. The CASI image after the envelope line is removed; Step 2: perform a subset operation on the CASI image after the envelope line removal described in step 1, to obtain an image subset with the spectral band at 540nm; The image subset at 540 nm is subjected to median filtering and pixel classification; step 4: analyzing and judging the pixel classification result.

Description

Translated fromChinese

一种利用CASI图像区别不同岩石中赤铁矿的方法A method for distinguishing hematite in different rocks using CASI images

技术领域technical field

本发明属于高光谱遥感矿物填图技术领域，具体涉及一种利用CASI图像区别不同岩石中赤铁矿的方法。The invention belongs to the technical field of hyperspectral remote sensing mineral mapping, and in particular relates to a method for distinguishing hematite in different rocks by using CASI images.

背景技术Background technique

CASI是一种重要的航空可见光－近红外遥感影像数据(遥感图像),波段范围400-950nm。赤铁矿是含三价铁离子的矿物，代表了相对氧化的环境。实际观察表明，许多矿床在空间上往往与赤铁矿有关，但赤铁矿可形成于不同的岩石中，代表不同的成矿环境。当前提取赤铁矿的方法包括：主成分分析法、吸收深度法、波段比值法等。上述方法虽然都可以提取赤铁矿信息，但无法判断赤铁矿形成于何种岩石中。因此，亟需研究一种利用CASI图像区别不同岩石中赤铁矿的方法。CASI is an important aerial visible light-near-infrared remote sensing image data (remote sensing image) with a wavelength range of 400-950 nm. Hematite is a mineral containing ferric ions and represents a relatively oxidizing environment. Practical observations show that many deposits tend to be spatially related to hematite, but hematite can form in different rocks and represent different metallogenic environments. The current methods of extracting hematite include: principal component analysis method, absorption depth method, band ratio method, etc. Although the above methods can extract hematite information, it is impossible to determine what kind of rock hematite is formed in. Therefore, there is an urgent need to develop a method for distinguishing hematite in different rocks using CASI images.

发明内容SUMMARY OF THE INVENTION

本发明针对现有技术不足提供一种利用CASI图像区别不同岩石中赤铁矿的方法，用于解决现有技术无法判断赤铁矿形成于何种岩石中的技术问题。Aiming at the deficiencies of the prior art, the present invention provides a method for distinguishing hematite in different rocks by using CASI images, so as to solve the technical problem that the prior art cannot judge which rock the hematite is formed in.

实现本发明目的的技术方案：The technical scheme that realizes the object of the present invention:

一种利用CASI图像区别不同岩石中赤铁矿的方法，包括以下步骤：A method for distinguishing hematite in different rocks using CASI images, including the following steps:

步骤一：预处理CASI图像，对预处理CASI图像进行包络线去除，获得包络线去除后的CASI图像；Step 1: Preprocess the CASI image, remove the envelope of the preprocessed CASI image, and obtain the CASI image after the envelope is removed;

步骤二：对步骤一中所述的包络线去除后的CASI图像进行取子集操作，获得光谱波段在540nm处的图像子集；Step 2: perform a subset operation on the CASI image after the envelope line removal described in step 1, to obtain a subset of images with a spectral band at 540 nm;

步骤三：对步骤二中所述的540nm处的图像子集进行中值滤波和像元分类；Step 3: Perform median filtering and pixel classification on the image subset at 540 nm described in Step 2;

步骤四：分析判断像元分类结果。Step 4: Analyze and judge pixel classification results.

如上所述步骤三：对步骤二中所述的540nm处的图像子集进行中值滤波和像元分类，包括：将540nm处的图像子集进行3×3中值滤波，将滤波结果中小于0.8的作为第一类像元，将滤波结果中介于0.8至0.96的作为第二类像元，将滤波结果中介于0.96至0.985的作为第三类像元，将滤波结果中大于0.985的作为第四类像元；Step 3 as described above: perform median filtering and pixel classification on the image subset at 540 nm described in step 2, including: performing 3×3 median filtering on the image subset at 540 nm, 0.8 is used as the first type of pixels, the filtering results are between 0.8 and 0.96 as the second type of pixels, the filtering results are between 0.96 and 0.985 as the third type of pixels, and the filtering results are greater than 0.985 as the first type of pixels. four types of pixels;

如上所述步骤四：分析判断像元分类结果，包括：Step 4 above: analyze and judge the pixel classification results, including:

A.将第二类像元和第三类像元在空间上彼此临近且第二、三类像元组成的区域中第二类像元占比超过10％的第二类像元判定为砂岩相关的赤铁矿像元，A. The second type of pixels where the second type of pixels and the third type of pixels are close to each other in space and the second type of pixels in the area composed of the second and third types of pixels accounts for more than 10% are judged as sandstone The associated hematite pixel,

B.将空间上只有第三类像元或相邻的第二、三类像元组成的区域中第二类像元占比不足10％的第三类像元判定为花岗岩相关的赤铁矿像元；B. Determining the third type of pixels with less than 10% of the second type of pixels in the area consisting of only the third type of pixels or the adjacent second and third type of pixels as granite-related hematite pixel;

C.将第一类和第四类像元判定为背景像元。C. Determine the first and fourth types of pixels as background pixels.

本发明的有益技术效果在于：The beneficial technical effect of the present invention is:

本发明设计的一种利用CASI图像区别不同岩石中赤铁矿的方法，操作步骤简单、数据处理速度快，且提取效果明显，便于不同岩石赤铁矿的识别。A method for distinguishing hematite in different rocks by using CASI images designed by the invention has simple operation steps, fast data processing speed, obvious extraction effect, and is convenient for the identification of hematite in different rocks.

具体实施方式Detailed ways

下面结合实施例对本发明作进一步详细说明。The present invention will be described in further detail below in conjunction with the embodiments.

一种利用CASI图像区别不同岩石中赤铁矿的方法，包括以下步骤：A method for distinguishing hematite in different rocks using CASI images, including the following steps:

步骤一：预处理CASI图像，对预处理CASI图像进行包络线去除，获得包络线去除后的CASI图像；Step 1: Preprocess the CASI image, remove the envelope of the preprocessed CASI image, and obtain the CASI image after the envelope is removed;

步骤二：对步骤一中所述的包络线去除后的CASI图像进行取子集操作，获得光谱波段在540nm处的图像子集；Step 2: perform a subset operation on the CASI image after the envelope line removal described in step 1, to obtain a subset of images with a spectral band at 540 nm;

步骤三：对步骤二中所述的540nm处的图像子集进行中值滤波和像元分类；Step 3: Perform median filtering and pixel classification on the image subset at 540 nm described in Step 2;

步骤四：分析判断像元分类结果。Step 4: Analyze and judge pixel classification results.

如上所述步骤三：对步骤二中所述的540nm处的图像子集进行中值滤波和像元分类，包括：将540nm处的图像子集进行3×3中值滤波，将滤波结果中小于0.8的作为第一类像元，将滤波结果中介于0.8至0.96的作为第二类像元，将滤波结果中介于0.96至0.985的作为第三类像元，将滤波结果中大于0.985的作为第四类像元；Step 3 as described above: perform median filtering and pixel classification on the image subset at 540 nm described in step 2, including: performing 3×3 median filtering on the image subset at 540 nm, 0.8 is used as the first type of pixels, the filtering results are between 0.8 and 0.96 as the second type of pixels, the filtering results are between 0.96 and 0.985 as the third type of pixels, and the filtering results are greater than 0.985 as the first type of pixels. four types of pixels;

如上所述步骤四：分析判断像元分类结果，包括：Step 4 above: analyze and judge the pixel classification results, including:

A.将第二类像元和第三类像元在空间上彼此临近且第二、三类像元组成的区域中第二类像元占比超过10％的第二类像元判定为砂岩相关的赤铁矿像元，A. The second type of pixels where the second type of pixels and the third type of pixels are close to each other in space and the second type of pixels in the area composed of the second and third types of pixels accounts for more than 10% are judged as sandstone The associated hematite pixel,

B.将空间上只有第三类像元或相邻的第二、三类像元组成的区域中第二类像元占比不足10％的第三类像元判定为花岗岩相关的赤铁矿像元；B. Determining the third type of pixels with less than 10% of the second type of pixels in the area consisting of only the third type of pixels or the adjacent second and third type of pixels as granite-related hematite pixel;

C.将第一类和第四类像元判定为背景像元。C. Determine the first and fourth types of pixels as background pixels.

上面结合实施例对本发明作了详细说明，但是本发明并不限于上述实施例，在本领域普通技术人员所具备的知识范围内，还可以在不脱离本发明宗旨的前提下作出各种变化。本发明中未作详细描述的内容均可以采用现有技术。The present invention has been described in detail above in conjunction with the embodiments, but the present invention is not limited to the above-mentioned embodiments, and within the scope of knowledge possessed by those of ordinary skill in the art, various changes can also be made without departing from the purpose of the present invention. The content that is not described in detail in the present invention can adopt the prior art.

Claims (1)

1. A method for distinguishing hematite ore in different rocks by using CASI images, comprising the steps of:

the method comprises the following steps: preprocessing a CASI image, and performing envelope elimination on the preprocessed CASI image to obtain a CASI image after envelope elimination;

step two: performing subset selection operation on the CASI image subjected to envelope removal in the step one to obtain an image subset of a spectral band at 540 nm;

step three: performing median filtering and pixel classification on the image subset at 540nm in the second step, including: 3 x 3 median filtering is carried out on the image subset at 540nm, pixels smaller than 0.8 in the filtering result are used as first type pixels, pixels between 0.8 and 0.96 in the filtering result are used as second type pixels, pixels between 0.96 and 0.985 in the filtering result are used as third type pixels, and pixels larger than 0.985 in the filtering result are used as fourth type pixels;

step four: analyzing and judging a pixel classification result; the method comprises the following steps:

A. the second type picture element with the second type picture element accounting for more than 10 percent of the area formed by the second and third type picture elements and the second type picture element close to each other in space is judged as the hematite picture element related to the sandstone,

B. judging the third type of picture element with the second type of picture element ratio less than 10 percent in the area which only has the third type of picture element or is formed by the adjacent second and third type of picture elements in space as the hematite picture element related to granite;

C. and judging the first type and the fourth type of pixels as background pixels.