A method of bloodstone in different rocks is distinguished using CASI imageTechnical field
The invention belongs to high-spectrum remote-sensing mineral charting technical fields, and in particular to a kind of to distinguish difference using CASI imageThe method of bloodstone in rock.
Background technique
CASI is a kind of important aviation visible light-near infrared range remote sensing image data (remote sensing images), wavelength band 400-950nm.Bloodstone is the mineral containing ferric ion, represents the environment of opposite oxidation.Actual observation shows that many mineral deposits existIt is spatially often related with bloodstone, but bloodstone can be formed in different rocks, represent different ore-forming settings.Work as premiseThe method for taking bloodstone includes: Principal Component Analysis, absorption depth method, band ratio method etc..Although the above method can mentionBloodstone information is taken, but can not judge which kind of rock bloodstone is formed in.Therefore, it needs to study a kind of utilization CASI image districtThe method of bloodstone in not different rocks.
Summary of the invention
The present invention provides a kind of method for distinguishing bloodstone in different rocks using CASI image for prior art deficiency,The technical problem that can not judge which kind of rock bloodstone be formed in for solving the prior art.
Realize the technical solution of the object of the invention:
A method of bloodstone in different rocks is distinguished using CASI image, comprising the following steps:
Step 1: pretreatment CASI image carries out envelope removal to pretreatment CASI image, after obtaining envelope removalCASI image；
Step 2: the CASI image after envelope described in step 1 removal is carried out taking subset op, obtains spectrumImage subset of the wave band at 540nm；
Step 3: median filtering is carried out to the image subset at 540nm described in step 2 and pixel is classified；
Step 4: pixel classification results are analyzed and determined.
Step 3 as described above: median filtering and pixel point are carried out to the image subset at 540nm described in step 2Class, comprising: the image subset at 540nm is subjected to 3 × 3 median filterings, by the conduct first kind picture in filter result less than 0.8Member, by conduct the second class pixel in filter result between 0.8 to 0.96, by the work in filter result between 0.96 to 0.985For third class pixel, 0.985 the 4th class pixel of conduct will be greater than in filter result；
Step 4 as described above: pixel classification results are analyzed and determined, comprising:
A. by the second class pixel and third class pixel be spatially in close proximity to one another and the region of second and third class pixel composition inSecond class pixel accounting is more than that 10% the second class pixel is determined as the relevant bloodstone pixel of sandstone,
B. the second class pixel in the region for spatially there was only third class pixel or second and third adjacent class pixel composition is accounted forIt is determined as the relevant bloodstone pixel of granite than the third class pixel less than 10%；
C. the first kind and the 4th class pixel are determined as backdrop pels.
The beneficial technical effect of the present invention lies in:
A kind of method for distinguishing bloodstone in different rocks using CASI image that the present invention designs, operating procedure is simple,Data processing speed is fast, and extraction effect is obvious, convenient for the identification of different rock bloodstone.
Specific embodiment
Below with reference to embodiment, invention is further described in detail.
A method of bloodstone in different rocks is distinguished using CASI image, comprising the following steps:
Step 1: pretreatment CASI image carries out envelope removal to pretreatment CASI image, after obtaining envelope removalCASI image；
Step 2: the CASI image after envelope described in step 1 removal is carried out taking subset op, obtains spectrumImage subset of the wave band at 540nm；
Step 3: median filtering is carried out to the image subset at 540nm described in step 2 and pixel is classified；
Step 4: pixel classification results are analyzed and determined.
Step 3 as described above: median filtering and pixel point are carried out to the image subset at 540nm described in step 2Class, comprising: the image subset at 540nm is subjected to 3 × 3 median filterings, by the conduct first kind picture in filter result less than 0.8Member, by conduct the second class pixel in filter result between 0.8 to 0.96, by the work in filter result between 0.96 to 0.985For third class pixel, 0.985 the 4th class pixel of conduct will be greater than in filter result；
Step 4 as described above: pixel classification results are analyzed and determined, comprising:
A. by the second class pixel and third class pixel be spatially in close proximity to one another and the region of second and third class pixel composition inSecond class pixel accounting is more than that 10% the second class pixel is determined as the relevant bloodstone pixel of sandstone,
B. the second class pixel in the region for spatially there was only third class pixel or second and third adjacent class pixel composition is accounted forIt is determined as the relevant bloodstone pixel of granite than the third class pixel less than 10%；
C. the first kind and the 4th class pixel are determined as backdrop pels.
The present invention is explained in detail above in conjunction with embodiment, but the present invention is not limited to above-described embodiments, at thisField those of ordinary skill within the scope of knowledge, can also make various changes without departing from the purpose of the present inventionChange.The content being not described in detail in the present invention can use the prior art.