CN116977871A - Stripe noise detection method and equipment and computer storage medium - Google Patents
Stripe noise detection method and equipment and computer storage medium Download PDFInfo
- Publication number
- CN116977871A CN116977871A CN202310812960.7A CN202310812960A CN116977871A CN 116977871 A CN116977871 A CN 116977871A CN 202310812960 A CN202310812960 A CN 202310812960A CN 116977871 A CN116977871 A CN 116977871A
- Authority
- CN
- China
- Prior art keywords
- remote sensing
- column
- pixels
- sensing image
- mean value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000001514 detection method Methods 0.000 title claims abstract description 127
- 238000000034 method Methods 0.000 claims abstract description 38
- 238000012545 processing Methods 0.000 claims abstract description 25
- 238000009499 grossing Methods 0.000 claims abstract description 14
- 230000008569 process Effects 0.000 claims description 15
- 238000002310 reflectometry Methods 0.000 claims description 11
- 238000012935 Averaging Methods 0.000 claims description 5
- 238000007781 pre-processing Methods 0.000 claims description 5
- 238000007689 inspection Methods 0.000 claims description 4
- 238000004891 communication Methods 0.000 claims description 3
- 238000004590 computer program Methods 0.000 claims description 3
- 238000004364 calculation method Methods 0.000 abstract description 10
- 230000002159 abnormal effect Effects 0.000 description 26
- 238000003384 imaging method Methods 0.000 description 20
- 238000010586 diagram Methods 0.000 description 18
- 238000012216 screening Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 6
- 239000000654 additive Substances 0.000 description 5
- 230000000996 additive effect Effects 0.000 description 5
- 230000005855 radiation Effects 0.000 description 4
- 230000003068 static effect Effects 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 3
- 238000003745 diagnosis Methods 0.000 description 3
- 230000001788 irregular Effects 0.000 description 3
- 238000012544 monitoring process Methods 0.000 description 3
- 238000011160 research Methods 0.000 description 3
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 230000014509 gene expression Effects 0.000 description 2
- 238000001228 spectrum Methods 0.000 description 2
- 238000007619 statistical method Methods 0.000 description 2
- 230000032683 aging Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000006731 degradation reaction Methods 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000008520 organization Effects 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 230000004044 response Effects 0.000 description 1
- 238000013519 translation Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/10—Terrestrial scenes
- G06V20/13—Satellite images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/20—Image preprocessing
- G06V10/30—Noise filtering
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Astronomy & Astrophysics (AREA)
- Remote Sensing (AREA)
- Image Processing (AREA)
Abstract
The application discloses a stripe noise detection method and equipment and a computer storage medium, and relates to the technical field of remote sensing image processing, wherein the stripe noise detection method is based on a wind cloud No. four B star rapid imager and comprises the following steps: superposing satellite remote sensing images within a certain time, and calculating the average value of pixel values of each row of pixels of the superposed remote sensing images to obtain original average value data; smoothing the original mean value data by using a window to obtain fitting mean value data; calculating the difference between the fitting mean value and the original mean value of each column of pixels, and differentiating the difference; setting a threshold value, comparing the threshold value with the differential value column by column, and if the differential value of the pixels in the column is larger than the threshold value, generating stripe noise in the column. The method has the advantages that the strip noise in the satellite image can be detected in real time in batches through simple steps, and the specific position of the strip noise in the satellite image can be detected rapidly and accurately, so that the detection of the strip noise is more efficient, and meanwhile, the workload of calculation and detection is reduced greatly.
Description
Technical Field
The application relates to the technical field of remote sensing image processing, in particular to a stripe noise detection method and equipment and a computer storage medium.
Background
The rapid imager is one of main loads carried in a wind cloud No. B star, and can realize rapid imaging of an area with the frequency of 2000km multiplied by 2000km for 1 minute.
In the remote sensing imaging process, various noises are inevitably generated on the acquired remote sensing image data, wherein stripe noises are typical, linear strips with fixed distribution directions are displayed on the remote sensing image, so that the image observation effect of the satellite remote sensing image acquired by the wind cloud No. four B star rapid imager is affected, and the subsequent processing problem is caused.
In recent years, there have been many studies on a method for removing abnormal bands from satellite remote sensing images, mainly including a digital filtering method, an image gray information statistical method, a variance method, and the like. However, in the current research of stripe noise, the position information of the stripe needs to be obtained in advance, and when the distribution of stripe noise does not have any rule, the manual interpretation mode is used for marking the position information, so that a great deal of manpower resources are consumed, and the limitation is particularly obvious when the sparse distribution of complex stripe noise is processed.
The existing patents related to stripe noise detection in images have some defects, for example, in the patent application of the self-detection and removal method (application number 200510027528.9) for spectrum domain noise of aviation hyperspectral remote sensing images, although the reflectance of the images can be used for detecting the spectrum domain noise of hyperspectral images, the calculation process is complex; for example, in the patent application "an improved algorithm for removing cross stripe noise in linear array image based on line tracking" (application number 201610649429.2), although a method for determining position information of cross stripe noise through connection between feature points is disclosed, the process of determining the noise position needs to be repeated for multiple times for comparison and judgment, which is complex as a whole; for another example, in the patent application "a method and a system for detecting radiation quality of a multi-source satellite remote sensing image product" (application number: 202210586657.5), although it is mentioned that a stripe detection can be performed on the multi-source satellite remote sensing image product, a frame to be detected of an image needs to be extracted during the stripe detection, and the detection is not performed on the whole image.
Disclosure of Invention
In view of the above, the present application provides a method, an apparatus, and a computer storage medium for detecting stripe noise, which can detect stripe noise in a satellite image in real time in batch through simple steps, and rapidly and accurately detect a specific position of the stripe noise in the satellite image, so that the detection of the stripe noise is more efficient, and meanwhile, the workload of calculation and detection is greatly reduced.
In a first aspect, the present application provides a stripe noise detection method, where the stripe noise detection method is based on a fast wind-cloud-type B star imager for stripe noise detection, and the fast imager is used for acquiring satellite remote sensing images in real time, and the stripe noise detection method includes:
dividing satellite remote sensing images according to time, and superposing the satellite remote sensing images within a certain time to obtain a superposed remote sensing image;
calculating the average value of pixel values of each column of pixels of the superimposed remote sensing image to obtain original average value data;
smoothing the original mean value data by using a window with a preset size to obtain fitting mean value data, wherein the fitting mean value data is a set of fitting mean values of each column of pixels of the superimposed remote sensing image;
calculating the difference between the fitting average value and the original average value of each column of pixels of the superimposed remote sensing image, and carrying out differential processing on the difference;
setting a threshold value, comparing the threshold value with the differential result column by column, and if the differential result of the pixels in the column is greater than the threshold value, generating stripe noise in the column; if the difference result of the pixels in the column is less than or equal to the threshold value, no stripe noise exists in the column;
calculating the average value of pixel values of each column of pixels of the superimposed remote sensing image, wherein obtaining the original average value data comprises the following steps:
summing the pixel values of each column of pixels of the superimposed remote sensing image and then averaging to obtain an original average value of each column of pixels;
arranging the original mean value of each column of pixels into rows according to the position sequence of the pixels in the corresponding column to obtain original mean value data;
the window has a center point, the window with a preset size is used for carrying out smoothing processing on the original mean value data to obtain fitting mean value data, the fitting mean value data is a set of fitting mean values of each column of pixels of the superimposed remote sensing image, and the method comprises the following steps:
and the center point of the window moves on the original mean value data row by row along the row direction, in the moving process, the original mean value of the overlapping part of the window and the original mean value data is fitted, the fitted value is used as the fitting mean value of the pixels of the column where the center point of the superimposed remote sensing image is located, and the set of the fitting mean values of the pixels of each column of the superimposed remote sensing image is the fitting mean value data.
Optionally, wherein:
the preset size of the window is 11.
Optionally, wherein:
the remote sensing image obtained by the rapid imager comprises a visible light range image, a near infrared range image and an infrared range image, when the remote sensing image is the visible light range image and the near infrared range image, the threshold value is set to be 0.03, and when the remote sensing image is the infrared range image, the threshold value is set to be 0.2.
Optionally, wherein:
before the satellite remote sensing images are divided according to time and the satellite remote sensing images within one hour are overlapped to obtain the overlapped remote sensing images, the stripe noise detection method further comprises the following steps:
the satellite remote sensing image obtained by the rapid imager is preprocessed, wherein the preprocessing comprises quality inspection, geographic positioning, radiometric calibration processing and conversion of pixel values of pixels in the visible light range image and the infrared range image into reflectivity according to the following formula:
ρ λ ＝SCALE×DN+OFFSET
wherein: ρ λ For the preprocessed satellite remote sensing image, SCALE provides a slope for converting the pixel value into reflectivity, DN provides an intercept for converting the pixel value into reflectivity for the satellite remote sensing image.
In a second aspect, the present application also provides a stripe noise detection device, comprising: a processor and a communication interface coupled to the processor for executing a computer program or instructions to implement the stripe noise detection method described in the first aspect.
In a third aspect, the present application also provides a computer storage medium having instructions stored therein which, when executed, implement the stripe noise detection method described in the first aspect.
Compared with the prior art, the stripe noise detection method, the stripe noise detection equipment and the computer storage medium provided by the application have the advantages that at least the following effects are realized:
according to the stripe noise detection method provided by the application, subsequent processing and detection are performed on the basis of the superimposed remote sensing images obtained by superposition, so that the accuracy of detection results is ensured. Then, pixel value average values of the superimposed remote sensing images are calculated along the column direction, average fitting is carried out on the obtained average value data by utilizing a window with a preset size, difference processing is carried out on the fitted average value of each column of pixels and the original average value, difference is carried out on the difference result, so that irregular fluctuation of the data is reduced, and abnormal data display is enhanced; and finally, screening abnormal strips through a set threshold value, so as to realize the rapid detection of the specific position of the strip noise in the wind cloud No. four B star rapid imager. Therefore, the application well combines the scanning imaging characteristics of the wind cloud No. B star rapid imager with the characteristics of the strips in the satellite remote sensing image, establishes a reliable strip noise detection method, and greatly improves the real-time on-orbit detection and diagnosis capability of the strips in the wind cloud No. B star remote sensing image; meanwhile, real-time batch detection of abnormal strips in the satellite remote sensing image can be realized, the workload of calculation and detection is greatly reduced, and the detection efficiency is further improved; in addition, the stripe noise detection method provided by the application can accurately identify and mark the single stripe, and the difference result can reflect the intensity of stripe noise, thereby providing assistance for anomaly detection and fault judgment.
Of course, it is not necessary for any one product embodying the application to achieve all of the technical effects described above at the same time.
Other features of the present application and its advantages will become apparent from the following detailed description of exemplary embodiments of the application, which proceeds with reference to the accompanying drawings.
Drawings
The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments of the application and together with the description, serve to explain the principles of the application.
Fig. 1 is a schematic flow chart of a stripe noise detection method according to an embodiment of the present application;
fig. 2 shows a banded satellite remote sensing image of a visible light channel 1 of a Fengyun No. B star rapid imager provided by an embodiment of the application;
FIG. 3 is a schematic diagram of noise components according to an ideal stripe noise model according to an embodiment of the present application;
fig. 4 shows a satellite remote sensing image of a visible light channel 1 of a fast imager of a B star B cloud with a noise component established according to an ideal stripe noise model according to an embodiment of the present application;
FIG. 5 is a schematic diagram showing an original average value of each row of pixels obtained after the stripe noise detection of the satellite remote sensing image shown in FIG. 4;
FIG. 6 is a schematic diagram showing a fitting average value of each column of pixels obtained after the stripe noise detection of the satellite remote sensing image shown in FIG. 4;
FIG. 7 is a schematic diagram showing a difference between a fitting average value and an original average value of each column of pixels obtained after performing stripe noise detection on the satellite remote sensing image shown in FIG. 4;
FIG. 8 is a schematic diagram showing a difference result of the difference value of each column of pixels obtained after the stripe noise detection of the satellite remote sensing image shown in FIG. 4;
FIG. 9 is a schematic diagram illustrating a threshold screening result obtained after the stripe noise detection of the satellite remote sensing image shown in FIG. 4;
fig. 10 is a detection step chart of performing stripe noise detection on satellite remote sensing images within 7 hours of a visible light channel 1 acquired by a fast wind-cloud star B imager at 9 and 14 days 2021 in an embodiment of the present application;
FIGS. 11 (a) and 11 (B) show two satellite remote sensing images randomly selected from all satellite remote sensing images within 7 hours of a visible light channel 1 acquired by a Fengyun No. four B star fast imager on day 2021, 9 and 14 in an embodiment of the present application;
fig. 12 is a schematic diagram of a detection result obtained by performing stripe noise detection on satellite remote sensing images in 7 hours of an infrared channel 7 acquired by a fast wind-cloud star B imager in 2021, 9 and 14 days in an embodiment of the present application;
fig. 13 (a) and 13 (B) show two satellite remote sensing images randomly selected from all satellite remote sensing images in 7 hours of an infrared channel 7 acquired from a fast imaging apparatus of a weather four-star B-star in 2021, 9 months and 14 days in an embodiment of the present application.
Detailed Description
Various exemplary embodiments of the present application will now be described in detail with reference to the accompanying drawings. It should be noted that: the relative arrangement of the components and steps, numerical expressions and numerical values set forth in these embodiments do not limit the scope of the present application unless it is specifically stated otherwise.
The following description of at least one exemplary embodiment is merely exemplary in nature and is in no way intended to limit the application, its application, or uses.
Techniques, methods, and apparatus known to one of ordinary skill in the relevant art may not be discussed in detail, but are intended to be part of the specification where appropriate.
In all examples shown and discussed herein, any specific values should be construed as merely illustrative, and not a limitation. Thus, other examples of exemplary embodiments may have different values.
It should be noted that: like reference numerals and letters denote like items in the following figures, and thus once an item is defined in one figure, no further discussion thereof is necessary in subsequent figures.
And 2021, 6 and 3 days, and the Fengyun No. four B satellites are successfully transmitted in the Wenchang satellite transmitting center. The wind cloud No. B star is a first business meteorological satellite of a wind cloud No. four series of a second generation static orbit meteorological satellite in China, and forms a static meteorological satellite business observation network together with the wind cloud No. A star, the wind cloud No. H star, the wind cloud No. G star and the wind cloud No. F star, realizes the business of the wind cloud No. four satellite, gradually replaces the wind cloud No. two, and becomes the dominant force of a world meteorological organization global meteorological satellite observation system.
Wind cloud number four B star carries four main loads: advanced stationary orbit radiation imagers (Advanced Geostationary Radiation Imager, AGRI), stationary orbit interferometric infrared detectors (Geostationary Interferometric Infrared Sounder, GIIRS), fast-speed imagers (GHI), and space environment monitoring packages (Space Environment Monitoring Instrument Package, SEP).
The rapid imager (GHI) is a similar load carried on a static orbit for the first time internationally, can realize rapid imaging of the region frequency of 2000km multiplied by 2000km for 1 minute, has the mesh spatial resolution of 250m, is the highest resolution of a global static orbit meteorological satellite, and further improves the continuous, flexible and high-resolution observation capability of typhoons, storm and mesoscale disastrous weather.
In the process that the wind cloud No. B star flies forwards along the track, the carried imaging instrument moves along with the wind cloud No. B star, after electromagnetic wave radiation is collected according to the working principle of the wind cloud No. B star GHI, a Charge-coupled Device (CCD) linear array detector can perform progressive scanning imaging, and when the offset and the relative gain value of the detector pixels are unequal, stripe noise can be generated, so that the distribution of the stripe noise is necessarily consistent with the scanning direction and is in the form of image rows or columns.
Currently, remote sensing imaging systems mainly comprise two different types of push-broom imaging systems and stripe-containing cross-track imaging devices. In the imaging process, the imaging sensor inevitably generates various noises on the acquired remote sensing image data, wherein stripe noises are typical, and the main reasons for the generation of the stripe noises are uneven responsivity of a detection unit caused by aging of a satellite CCD (charge coupled device) detector, and vignetting of an imaging optical system, so that the image is represented as linear noises with fixed directions. Under the influence of the influence, the image observation effect of the cloud No. four B star GHI image is influenced, and the subsequent processing problems such as classification, target detection, quantitative application and the like can be caused. Because the response of each pixel in the image to noise has no correlation, the stripe noise can show strong randomness and unsupervised property on the remote sensing image, and is difficult to remove by using a general calibration method.
In recent years, there have been many studies on an abnormal band removal algorithm on satellite remote sensing images, mainly including a digital filtering method, an image gray information statistical method, a variance method, and the like, but there have been few studies on abnormal band detection on satellite remote sensing images. At present, the research on the stripe noise often needs to acquire the position information of the stripe in advance. In the prior art, the position information of the stripe noise is often obtained by using a manual translation mode, but when the distribution of the stripe noise does not have any rule, a great deal of manpower resources are consumed by marking the position information by using a manual interpretation mode, and the limitation is particularly obvious when the complex stripe noise with sparse distribution is processed.
In order to solve the technical problems, the application provides a stripe noise detection method, a stripe noise detection device and a computer storage medium, which can detect stripe noise in satellite images in real time in batches through simple steps, and rapidly and accurately detect specific positions of the stripe noise in the satellite images, so that the detection of the stripe noise is more efficient, and meanwhile, the workload of calculation and detection is greatly reduced.
The following detailed description refers to the accompanying drawings and specific embodiments.
Fig. 1 is a flow chart of a stripe noise detection method according to an embodiment of the present application.
As shown in fig. 1, the embodiment of the application provides a stripe noise detection method, which is based on a wind-cloud-number-four-B-star rapid imager for stripe noise detection, wherein the rapid imager is used for acquiring satellite remote sensing images in real time, and the stripe noise detection method comprises the following steps:
s100, dividing satellite remote sensing images according to time, and superposing the satellite remote sensing images within a certain time to obtain a superposed remote sensing image;
s200, calculating the average value of pixel values of each row of pixels of the superimposed remote sensing image to obtain original average value data; calculating the average value of pixel values of each column of pixels of the superimposed remote sensing image, wherein obtaining the original average value data comprises the following steps:
summing the pixel values of each column of pixels of the superimposed remote sensing image and then averaging to obtain an original average value of each column of pixels;
arranging the original mean value of each column of pixels into rows according to the position sequence of the pixels in the corresponding column to obtain original mean value data;
s300, performing smoothing on the original mean value data by using a window with a preset size to obtain fitting mean value data, wherein the fitting mean value data is a set of fitting mean values of each column of pixels of the superimposed remote sensing image; the window has a center point, the window with a preset size is used for carrying out smoothing processing on the original mean value data to obtain fitting mean value data, the fitting mean value data is a set of fitting mean values of each column of pixels of the superimposed remote sensing image, and the method comprises the following steps:
the center point of the window moves on the original mean value data row by row along the row direction, in the moving process, the original mean value of the overlapping part of the window and the original mean value data is fitted, the fitted value is used as the fitting mean value of the pixels of the column where the center point of the superimposed remote sensing image is located, and the set of the fitting mean values of the pixels of each column of the superimposed remote sensing image is the fitting mean value data;
s400, calculating the difference between the fitting mean value and the original mean value of each column of pixels of the superimposed remote sensing image, and carrying out differential processing on the difference;
s500, setting a threshold value, comparing the threshold value with the differential result column by column, and if the differential result of the pixels in the column is larger than the threshold value, generating stripe noise in the column; if the differential result of the pixels in the column is less than or equal to the threshold, then the column has no stripe noise.
Based on the above, in the stripe noise detection method provided by the embodiment of the application, a plurality of satellite remote sensing images acquired by a rapid imager within a certain time are superposed, and subsequent processing and detection are performed on the basis of the superposed remote sensing images obtained by superposition, so that the accuracy of detection results is ensured. Then, pixel value average values of the superimposed remote sensing images are calculated along the column direction, the calculated pixel value average values are arranged in rows column by column, average fitting is carried out on the obtained average value data by utilizing a window with a preset size, and fitting average value data after smoothing processing is obtained; performing difference processing on the fitting average value and the original average value of each column of pixels, and differentiating the difference result to reduce irregular fluctuation of data and enhance abnormal data display; and finally, comparing the differential results column by column through a set threshold value, and screening abnormal strips. Because the pixels at the ground object change position in the satellite remote sensing image are generally mixed pixels, the pixel values of adjacent pixels only gradually change and are not suddenly changed, so that whether the pixel values of the adjacent pixels are suddenly changed or not can be judged by utilizing a difference result, and whether the stripes exist or not can be further judged. If the difference result is larger than the threshold value, the pixel values of the adjacent pixels are suddenly changed, and further the existence of stripe noise in the column is indicated, so that the rapid detection of the specific position of the stripe noise in the wind cloud four-star B rapid imaging instrument is realized. Therefore, the embodiment of the application well combines the scanning imaging characteristics of the wind cloud No. B star rapid imager with the characteristics of the strips in the satellite remote sensing image, establishes a reliable strip noise detection method, has simple steps and high efficiency, and greatly improves the real-time on-orbit detection and diagnosis capability of the strips in the wind cloud No. B star remote sensing image; meanwhile, the stripe noise detection method provided by the embodiment of the application is based on the image data acquired by the wind-cloud-number-four-B-star rapid imager, considers the space and time statistical characteristics of a plurality of remote sensing images in continuous time, can realize real-time batch detection of abnormal stripes in satellite remote sensing images, greatly reduces the workload of calculation and detection, and further improves the detection efficiency; in addition, the stripe noise detection method provided by the embodiment of the application not only can realize accurate detection of the abnormal stripe position in the satellite remote sensing image, but also can accurately identify and mark a single stripe, and the difference result can reflect the intensity of stripe noise, thereby providing assistance for abnormal detection and fault judgment.
It should be noted that, in the embodiment of the present application, the pixel value of the pixel may be understood as the gray value of the pixel.
In some examples, the embodiment of the application takes an abnormal stripe in a satellite remote sensing image obtained by a wind cloud No. B star rapid imager as column additive noise, and the stripe noise at the moment is distributed in the satellite remote sensing image along the column direction as an example for illustration, and is not particularly limited.
In some examples, satellite remote sensing images within one hour may be superimposed in dividing the satellite remote sensing images in time and superimposing the satellite remote sensing images over a period of time to obtain superimposed remote sensing images. In consideration of the fact that abnormal stripes in satellite remote sensing images are mainly caused by imperfect calibration, the stripe noise detection method usually lasts for a period of time and even months on a detector, therefore, the satellite remote sensing images acquired by a rapid imager along time can be divided in an hour unit, image data in one hour are processed each time, all image data in the hour are overlapped and summed, and an average value is obtained along a column direction instead of each image processing each time, and under the condition that accuracy of detection results is well maintained, workload of actual on-orbit satellite calculation and monitoring is greatly reduced.
In some examples, when the smoothing process is performed on the raw mean value data, a window with a preset size may be utilized, the raw mean value in the window is subjected to mean value fitting when moving on a data set formed by the raw mean values of pixels in each column, the fitted result is used as a fitted mean value of the position of the central point of the window, the fitted mean value is assigned to a new array, after the smoothing process is finished, the fitted mean values of pixels in all columns are arranged in columns, finally fitted mean value data is obtained, and whether abnormal stripes exist in the current column or not is conveniently detected according to the difference between the fitted mean value and the raw mean value. Therefore, the method for detecting the stripe noise provided by the embodiment of the application can realize the accurate detection of the abnormal stripe through simple steps, does not need to use complex calculation and complicated formulas, greatly reduces the workload of calculation and detection, and further improves the detection efficiency.
For example, since the embodiment of the present application uses stripe noise as column additive noise for illustration, the formed raw mean data is a data set with the raw mean of each column of pixels arranged in rows, and thus, the window selected in the embodiment of the present application is a window arranged in the row direction; if the stripe noise is a row additive noise, a window aligned in the column direction may be selected for the mean fit, which is not specifically illustrated herein.
Illustratively, the predetermined size of the window is 11. Specifically, the window is a window in which 11 pixel points are arranged one by one along the row direction, the center point at this time is located at the middle position of the window having 11 points, 5 pixel points are located on the left side of the center point, and 5 pixel points are located on the right side of the center point. In the smoothing process, the window starts to move along the row direction from the first original mean value of the original mean value data, specifically, the center point of the window starts to gradually move along the row direction from the original mean value corresponding to the first column of pixels, when the center point of the window is coincident with the original mean value position corresponding to a certain column of pixels, the original mean value of the overlapping part of 11 pixel points in the window and the original mean value data is averaged, and the value obtained by the averaging is assigned to the corresponding position of the column of pixels corresponding to the center point in a new array, namely, the fitting mean value of the column of pixels. When the preset size of the window is 11, a good smoothing effect can be realized just on the basis of keeping the key information, and if the preset size of the window is too large, the key information can be missed; if the preset size of the window is too small, the smoothing function is not achieved.
For example, in the smoothing process, if the position of the pixel point of the window is not completely coincident with the original mean value data, that is, the position of part of the pixel points of the window does not have the original mean value, the mirror image processing is required to be performed on the window. For example, when the center point of the window is coincident with the leftmost original mean value of the original mean value data, and at this time, the positions of the 5 pixel points on the left side of the center point do not have the original mean value, the original mean value corresponding to the 5 pixel points on the right side of the center point needs to be mirrored to the left side, and at this time, the left side and the right side of the center point are completely symmetrical.
In some examples, calculating the difference between the fitted mean value and the original mean value for each column of pixels of the overlaid remote sensing image refers to subtracting the original mean value from the fitted mean value for each column of pixels of the overlaid image.
In some examples, the remote sensing image acquired by the fast imager includes a visible range image, a near infrared range image, and an infrared range image, the threshold is set to 0.03 when the remote sensing image is a visible range image and a near infrared range image, and the threshold is set to 0.2 when the remote sensing image is an infrared range image.
Based on the above, when the differential result is set to a threshold value for abnormal band screening, the threshold value can be set according to experience in actual engineering, and specific set values of the threshold value are different due to a visible light channel, a near infrared channel and an infrared channel. In actual business and research, the threshold value can be set to be 0.03 to screen abnormal strips of a visible light channel and a near infrared channel, the threshold value can be set to be 0.2 to screen abnormal strips of an infrared channel, and if the difference result is larger than the set threshold value, the existence of strip noise in the array is judged, so that the abnormal strips affecting satellite remote sensing image data application can be detected very well, and the accuracy of strip noise detection is further improved. For example, the threshold may be reduced by a small amount to detect even smaller abnormal bands.
At present, most stripe noise detection and removal models are used for directly finding an ideal clean stripe-free image from an actually measured image. However, considering the good characteristics of the stripe noise, if modeling and constraint can be performed on the characteristics of the stripe noise, the estimation of the stripe component will be more accurate, and thus the obtained stripe detection result will be more ideal.
Therefore, in the on-orbit detection of the wind cloud satellite B stripe noise, the embodiment of the application combines the characteristics of the stripe noise in the satellite remote sensing image, presets the position and intensity information of the stripe, establishes an ideal stripe noise model on the non-stripe satellite remote sensing image of the visible light channel 1, and judges whether the detection result of the embodiment of the application has consistency with the stripe noise designed in the model, thereby proving the effectiveness of the embodiment of the application. The specific set-up procedure for the ideal stripe noise model is as follows.
Firstly, in the embodiment of the application, stripe effect in the images of the wind cloud No. four B star rapid imager is modeled as column additive noise, and then the degradation process of the non-stripe satellite remote sensing image is expressed as follows:
X+S＝Y
wherein, the image containing the stripes can be expressed as the sum of the original image and the stripe component, wherein Y represents the degenerated satellite remote sensing image, X represents the non-stripe satellite remote sensing image, and S represents the stripe component.
In order to more accurately design the fringe component to represent the stripe noise in the wind cloud number four B star fast imager in orbit, the key issue is to fully exploit the directionality and low rank and sparse features of the fringe component information and describe them with an appropriate model.
Therefore, when the embodiment of the application is used for describing the characteristic of sparse distribution of the strip components in the global scale, the sparsity of the strip components can be quantified by directly designing the following sparsity measurement formula:
wherein, for a solution vector, x= [ x ] 1 ，x 2 ，……x n ]And (3) representing the sparse coefficient, wherein N is the column pixel dimension of the satellite remote sensing image. The greater the value of Sparseness (x), the greater the sparsity of x, indicating that the stripes are more sparse.
After adding column additive noise in the range of 20% -50% of the image mean value in the banded satellite remote sensing image, observable bands appear in the image, wherein the image mean value refers to a result obtained by averaging pixel values of all pixels of the whole banded satellite remote sensing image, namely an image intensity mean value. Therefore, in order to make the subsequent observation of the stripe detection result more intuitive, the embodiment of the application establishes the following expression equation of the stripe from left to right by using the stripe noise from 20% of weak noise to 50% of strong noise component of the image intensity mean value:
y＝B i X i
wherein X is i Represents the ith column stripe vector, B i Represents the noise intensity of the ith column stripe, B i Satisfy (B) i -B i-1 ＝B i-1 -B i-2 )。
Fig. 2 shows a banded satellite remote sensing image of a visible light channel 1 of a Fengyun No. B star rapid imager provided by an embodiment of the application; FIG. 3 is a schematic diagram of noise components according to an ideal stripe noise model according to an embodiment of the present application; fig. 4 shows a satellite remote sensing image of a visible light channel 1 of a fast imager of a B star B cloud with a noise component established according to an ideal stripe noise model according to an embodiment of the present application; FIG. 5 is a schematic diagram showing an original average value of each row of pixels obtained after the stripe noise detection of the satellite remote sensing image shown in FIG. 4; FIG. 6 is a schematic diagram showing a fitting average value of each column of pixels obtained after the stripe noise detection of the satellite remote sensing image shown in FIG. 4; FIG. 7 is a schematic diagram showing a difference between a fitting average value and an original average value of each column of pixels obtained after performing stripe noise detection on the satellite remote sensing image shown in FIG. 4; FIG. 8 is a schematic diagram showing a difference result of the difference value of each column of pixels obtained after the stripe noise detection of the satellite remote sensing image shown in FIG. 4; fig. 9 is a schematic diagram of a threshold screening result obtained after the stripe noise detection of the satellite remote sensing image shown in fig. 4.
The method for detecting stripe noise in the visible light channel 1 according to the embodiment of the present application is used to detect stripe of the satellite remote sensing image of the visible light channel 1 including the ideal stripe noise model, as shown in fig. 5 to 9, by combining the stripe-free satellite remote sensing image of the visible light channel 1 shown in fig. 2 with the noise component according to the ideal stripe noise model shown in fig. 3, so as to obtain the stripe-free satellite remote sensing image of the visible light channel 1 including the noise component according to the ideal stripe noise model shown in fig. 4. In fig. 5 to 9, the horizontal axis represents the number of each row of pixels of the satellite remote sensing image, the vertical axis represents the magnitude of a numerical value obtained by processing the pixel value of a certain row of pixels, and I represents the magnitude.
The vertical axis in fig. 9 indicates a difference value obtained by differentiating the difference between the fitting average value and the original average value, and the larger the absolute value of the difference value is, the larger the pixel value jump of the column of pixels is, and the more the stripe noise is obvious. Referring to the threshold value screening result shown in fig. 9, the result obtained after the stripe noise detection method provided by the embodiment of the present application is matched with the stripe noise component established according to the ideal stripe noise model shown in fig. 3, which proves that the stripe noise detection method provided by the embodiment of the present application can detect the stripe noise more accurately, and the intensity of the stripe noise can be reflected by the magnitude of the differential value. Therefore, the method for detecting the stripe noise provided by the embodiment of the application processes the satellite remote sensing image in each hour, so that the abnormal row in the satellite remote sensing image in each hour can be obtained, and the stripe occurrence rate in the hour is obtained, is an index related to the scanning row of the imager, and can reflect the occurrence frequency and the quality of the image in the satellite remote sensing image obtained by the rapid imager.
In some examples, before dividing the satellite remote sensing image according to time and overlapping the satellite remote sensing image within one hour, the method for detecting the stripe noise further comprises:
the satellite remote sensing image obtained by the rapid imager is preprocessed, wherein the preprocessing comprises quality inspection, geographic positioning, radiometric calibration processing and conversion of pixel values of pixels in the visible light range image and the infrared range image into reflectivity according to the following formula:
ρ λ ＝SCALE×DN+OFFSET
wherein: ρ λ For the preprocessed satellite remote sensing image, SCALE provides a slope for converting the pixel value into reflectivity, DN provides an intercept for converting the pixel value into reflectivity for the satellite remote sensing image.
Based on the method, the obtained satellite remote sensing image of the wind cloud No. four B star rapid imager can be used as 0-level source packet data, the 0-level source packet data is preprocessed, and the method specifically comprises quality inspection, geographic positioning and radiometric calibration processing, and as pixel values do not have unit significance, digital quantized values (Digital numbers, DN for short) of the satellite remote sensing images of the visible light channel and the infrared channel can be converted into reflectivity and bright temperature in a mode of the formula or the lookup table, L1-level remote sensing image data is obtained, the L1-level remote sensing image data is used for carrying out subsequent stripe noise detection, dimensionless pixel values can be converted into reflectivity with practical physical significance, and errors caused by the sensor are eliminated.
Fig. 10 is a detection step chart of performing stripe noise detection on satellite remote sensing images within 7 hours of a visible light channel 1 acquired by a fast wind-cloud star B imager at 9 and 14 days 2021 in an embodiment of the present application; FIGS. 11 (a) and 11 (B) show two satellite remote sensing images randomly selected from all satellite remote sensing images within 7 hours of a visible light channel 1 acquired by a Fengyun No. four B star fast imager on day 2021, 9 and 14 in an embodiment of the present application; fig. 12 is a schematic diagram of a detection result obtained by performing stripe noise detection on satellite remote sensing images in 7 hours of an infrared channel 7 acquired by a fast wind-cloud star B imager in 2021, 9 and 14 days in an embodiment of the present application; fig. 13 (a) and 13 (B) show two satellite remote sensing images randomly selected from all satellite remote sensing images in 7 hours of an infrared channel 7 acquired by a fast imaging apparatus of a weather four-star B-star in 2021, 9 months and 14 days in an embodiment of the present application. The detection step chart shown in fig. 10 refers to an image formed by sequentially arranging an original mean value schematic diagram of each column of pixels, a fitting mean value schematic diagram of each column of pixels, a difference value schematic diagram of the fitting mean value and the original mean value of each column of pixels, and a difference result schematic diagram of the difference value of each column of pixels in the process of carrying out stripe noise detection on the satellite remote sensing image in 7 hours of the visible light channel 1 acquired by the wind cloud fast imaging device of the star B at the day 2021, 9 and 14; in fig. 10 and 12, the horizontal axis represents the number of each column of pixels of the satellite remote sensing image, and the vertical axis represents the magnitude of the value obtained by processing the pixel value of a certain column of pixels.
In order to further verify the effectiveness and accuracy of the stripe noise detection method provided by the embodiment of the application, as shown in fig. 10 to 13, satellite remote sensing images acquired by a wind cloud four-star B rapid imager on day 9 and 14 of 2021 are preprocessed to obtain L1-level remote sensing image data, and the preprocessing process is performed to obtain L1-level remote sensing image data, specifically L1-level B-level remote sensing image data, and abnormal stripe detection is performed on L1-level B-level remote sensing image data of a visible light channel 1 and an infrared channel according to the stripe noise detection method provided by the embodiment of the application. Referring to fig. 10 and 11, it can be seen that no stripe noise is detected in the satellite remote sensing image of the visible light channel 1 within 7 hours of the day 14 of 9 of 2021 for the fast imaging device of the star B of the wind cloud, and meanwhile stripe noise detection is performed on the visible light channels 1 to 6 of the day 14 of 9 of 2021 for the fast imaging device of the star B of the wind cloud, so that the occurrence rate of stripe noise of each visible light channel in the day is 0 and is consistent with the actual L1B-level remote sensing image data of the visible light channel; referring to fig. 12 and 13, it can be seen that, according to the actual experience of the infrared channel, the threshold is set to 0.2, and the band noise detection method provided by the embodiment of the application is used to detect the band of the infrared channel 7, it is found that the infrared channel 7 obtained by the wind cloud four-star B-star rapid imager at the 9 th month 14 of 2021 detects abnormal bands in each hour, and the number of the bands found in the 7 th hour is the largest and is consistent with the band noise condition in the actual satellite remote sensing image.
Based on the same inventive concept, the present application also provides a stripe noise detection apparatus comprising: the processor and the communication interface are coupled to the processor for executing the computer program or instructions to implement the stripe noise detection method described in the above embodiments.
Compared with the prior art, the beneficial effects of the stripe detection device are the same as those of the stripe noise detection method described in the above embodiment, and are not described here again.
Based on the same inventive concept, the present application also provides a computer storage medium having instructions stored therein, which when executed, implement the stripe noise detection method described in the above embodiments.
Compared with the prior art, the beneficial effects of the computer storage medium are the same as those of the stripe noise detection method described in the above embodiment, and are not described here again.
In summary, the stripe noise detection method, the stripe noise detection device and the computer storage medium provided by the application at least realize the following beneficial effects:
according to the stripe noise detection method provided by the application, subsequent processing and detection are performed on the basis of the superimposed remote sensing images obtained by superposition, so that the accuracy of detection results is ensured. Then, pixel value average values of the superimposed remote sensing images are calculated along the column direction, average fitting is carried out on the obtained average value data by utilizing a window with a preset size, difference processing is carried out on the fitted average value of each column of pixels and the original average value, difference is carried out on the difference result, so that irregular fluctuation of the data is reduced, and abnormal data display is enhanced; and finally, screening abnormal strips through a set threshold value, so as to realize the rapid detection of the specific position of the strip noise in the wind cloud No. four B star rapid imager. Therefore, the application well combines the scanning imaging characteristics of the wind cloud No. B star rapid imager with the characteristics of the strips in the satellite remote sensing image, establishes a reliable strip noise detection method, and greatly improves the real-time on-orbit detection and diagnosis capability of the strips in the wind cloud No. B star remote sensing image; meanwhile, real-time batch detection of abnormal strips in the satellite remote sensing image can be realized, the workload of calculation and detection is greatly reduced, and the detection efficiency is further improved; in addition, the stripe noise detection method provided by the application can accurately identify and mark the single stripe, and the difference result can reflect the intensity of stripe noise, thereby providing assistance for anomaly detection and fault judgment.
While certain specific embodiments of the application have been described in detail by way of example, it will be appreciated by those skilled in the art that the above examples are for illustration only and are not intended to limit the scope of the application. It will be appreciated by those skilled in the art that modifications may be made to the above embodiments without departing from the scope and spirit of the application. The scope of the application is defined by the appended claims.
Claims (6)
1. The stripe noise detection method is characterized by carrying out stripe noise detection based on a wind cloud No. four B star rapid imager, wherein the rapid imager is used for acquiring satellite remote sensing images in real time, and the stripe noise detection method comprises the following steps:
dividing the satellite remote sensing image according to time, and superposing the satellite remote sensing image within a certain time to obtain a superposed remote sensing image;
calculating the average value of pixel values of each row of pixels of the superimposed remote sensing image to obtain original average value data;
smoothing the original mean value data by using a window with a preset size to obtain fitting mean value data, wherein the fitting mean value data is a set of fitting mean values of each column of pixels of the superimposed remote sensing image;
calculating the difference between the fitting average value and the original average value of each column of pixels of the superimposed remote sensing image, and carrying out differential processing on the difference;
setting a threshold value, comparing the threshold value with the differential result column by column, and if the differential result of the pixels in the column is greater than the threshold value, generating stripe noise in the column; if the difference result of the pixels in the column is less than or equal to the threshold value, no stripe noise exists in the column;
the calculating the average value of the pixel values of each column of pixels of the superimposed remote sensing image, and obtaining the original average value data includes:
summing pixel values of each column of pixels of the superimposed remote sensing image and then averaging to obtain an original average value of each column of pixels;
arranging the original mean value of each column of pixels into rows according to the position sequence of the pixels in the corresponding column to obtain original mean value data;
the window has a center point, the window with a preset size is used for carrying out smoothing processing on the original mean value data to obtain fitting mean value data, and the fitting mean value data is a set of fitting mean values of each column of pixels of the superimposed remote sensing image and comprises:
and the central point of the window moves on the original mean value data row by row along the row direction, in the moving process, the original mean value of the overlapping part of the window and the original mean value data is fitted, the fitted value is used as the fitting mean value of the pixels of the central point of the superimposed remote sensing image, and the set of the fitting mean values of the pixels of each row of the superimposed remote sensing image is the fitting mean value data.
2. The banding noise detection method according to claim 1, wherein the preset size of the window is 11.
3. The method according to claim 1, wherein the remote sensing image obtained by the fast imager includes a visible light range image, a near infrared range image, and an infrared range image, and the threshold is set to 0.03 when the remote sensing image is a visible light range image and a near infrared range image, and is set to 0.2 when the remote sensing image is an infrared range image.
4. The banding noise detection method according to claim 3, wherein before said dividing said satellite remote sensing image by time and superimposing said satellite remote sensing image within one hour, said banding noise detection method further comprises:
preprocessing the satellite remote sensing image acquired by the rapid imager, wherein the preprocessing comprises quality inspection, geographic positioning, radiometric calibration processing and converting pixel values of pixels in the visible light range image and the infrared range image into reflectivity according to the following formula:
ρ λ ＝SCALE×DN+OFFSET
wherein: ρ λ For the preprocessed satellite remote sensing image, SCALE provides a slope for converting the pixel value into reflectivity, DN provides an intercept for converting the pixel value into reflectivity for the satellite remote sensing image.
5. A banding noise detection apparatus characterized by comprising: a processor and a communication interface coupled to the processor for running a computer program or instructions to implement the banding noise detection method according to any one of claims 1 to 4.
6. A computer storage medium having instructions stored therein which, when executed, implement the banding noise detection method of any one of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310812960.7A CN116977871A (en) | 2023-07-04 | 2023-07-04 | Stripe noise detection method and equipment and computer storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310812960.7A CN116977871A (en) | 2023-07-04 | 2023-07-04 | Stripe noise detection method and equipment and computer storage medium |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN116977871A true CN116977871A (en) | 2023-10-31 |
Family
ID=88477568
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310812960.7A Pending CN116977871A (en) | 2023-07-04 | 2023-07-04 | Stripe noise detection method and equipment and computer storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116977871A (en) |
-
2023
- 2023-07-04 CN CN202310812960.7A patent/CN116977871A/en active Pending
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Li et al. | An evaluation of the use of atmospheric and BRDF correction to standardize Landsat data | |
| Gladkova et al. | Quantitative restoration for MODIS band 6 on Aqua | |
| Cook et al. | IKONOS technical performance assessment | |
| CN112529807B (en) | Relative radiation correction method and device for satellite image | |
| CN106940887B (en) | GF-4 satellite sequence image cloud and cloud shadow detection method | |
| CN111337434A (en) | Mining area reclamation vegetation biomass estimation method and system | |
| Sterckx et al. | Atmospheric correction of APEX hyperspectral data | |
| CN113447137B (en) | Surface temperature inversion method for unmanned aerial vehicle broadband thermal imager | |
| Lu et al. | An evaluation of FY-3C satellite data quality at ECMWF and the Met Office | |
| CN113970376A (en) | Satellite infrared load calibration method based on ocean area reanalysis data | |
| US7796153B1 (en) | Equalization system and method for an imaging sensor | |
| CN110689505B (en) | Scene-based satellite-borne remote sensing instrument self-adaptive correction method and system | |
| CN116664947A (en) | Blue algae bloom monitoring method and system based on satellite observation data | |
| CN113049606A (en) | Large-area high-precision insulator pollution distribution assessment method | |
| Wang et al. | Removing temperature drift and temporal variation in thermal infrared images of a UAV uncooled thermal infrared imager | |
| CN116977871A (en) | Stripe noise detection method and equipment and computer storage medium | |
| CN111199557A (en) | Quantitative analysis method and system for decay of remote sensor | |
| Takahashi | Algorithm theoretical basis document (ATBD) for GSICS infrared inter-calibration of imagers on MTSAT-1R/-2 and himawari-8/-9 using AIRS and IASI hyperspectral observations | |
| Harris et al. | Radiometric homogenisation of aerial images by calibrating with satellite data | |
| Li et al. | Assimilation of Chinese Fengyun-3B Microwave Temperature Sounder radiances into the Global GRAPES system with an improved cloud detection threshold | |
| Reulke et al. | Image quality and image resolution | |
| KR20100028337A (en) | Method for calibration of coms throught cloud using cloud target | |
| CN110866942B (en) | On-orbit radiation response tracking method and system for multi-source target fusion | |
| Khetkeeree et al. | Satellite image restoration using adaptive high boost filter based on in-flight point spread function | |
| CN114677275B (en) | High-frequency repeated staring imaging space-based remote sensing load on-orbit testing method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |