CN116503274B - Image color homogenizing method and device based on image overlapping area - Google Patents
Image color homogenizing method and device based on image overlapping area Download PDFInfo
- Publication number
- CN116503274B CN116503274B CN202310372537.XA CN202310372537A CN116503274B CN 116503274 B CN116503274 B CN 116503274B CN 202310372537 A CN202310372537 A CN 202310372537A CN 116503274 B CN116503274 B CN 116503274B
- Authority
- CN
- China
- Prior art keywords
- image
- images
- radiation characteristic
- aerial
- belt
- 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.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 61
- 230000005855 radiation Effects 0.000 claims abstract description 90
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 claims abstract description 28
- 238000004364 calculation method Methods 0.000 claims description 24
- 230000008569 process Effects 0.000 claims description 21
- 238000002156 mixing Methods 0.000 claims description 14
- 238000007781 pre-processing Methods 0.000 claims description 8
- 238000012216 screening Methods 0.000 claims description 4
- 238000012163 sequencing technique Methods 0.000 claims description 4
- 238000000605 extraction Methods 0.000 claims description 2
- 230000000694 effects Effects 0.000 abstract description 12
- 238000010586 diagram Methods 0.000 description 10
- 238000000265 homogenisation Methods 0.000 description 6
- 238000001914 filtration Methods 0.000 description 3
- 238000003384 imaging method Methods 0.000 description 3
- 230000000873 masking effect Effects 0.000 description 3
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 239000003643 water by type Substances 0.000 description 2
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 239000003086 colorant Substances 0.000 description 1
- 238000004590 computer program Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 230000008030 elimination Effects 0.000 description 1
- 238000003379 elimination reaction Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000008439 repair process Effects 0.000 description 1
Classifications
-
- G06T5/92—
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformation in the plane of the image
- G06T3/40—Scaling the whole image or part thereof
- G06T3/4038—Scaling the whole image or part thereof for image mosaicing, i.e. plane images composed of plane sub-images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/90—Determination of colour characteristics
Abstract
The invention discloses an image homogenizing method and device based on an image overlapping area and a computer readable storage medium, wherein the method comprises the following steps: acquiring a plurality of preprocessed aerial belt images; extracting an overlapping region from adjacent aerial band images based on a preset shielding water area mask, and extracting a radiation characteristic curve of the adjacent aerial band images; and carrying out first color homogenizing treatment on the adjacent avionics images according to the pixels of the overlapping area and the radiation characteristic curve. According to the invention, a plurality of preprocessed aerial belt images can be obtained, the two aerial belt images are subjected to color homogenizing treatment according to pixel information of the overlapped area and radiation characteristic information of the aerial belt images to be subjected to color homogenizing on the basis of a preset mask of a shielding water area in an overlapped area of the two adjacent aerial belt images, so that deviation of color homogenizing is reduced, and a color homogenizing effect is improved.
Description
Technical Field
The present invention relates to the field of image color processing, and in particular, to an image color homogenizing method and apparatus based on an image overlapping region.
Background
The SAR image is an image acquired through earth observation by a synthetic aperture radar installed on a flight platform such as an airplane, a satellite, a spacecraft and the like, and can be used for carrying out earth observation and analysis all the time and all the weather. Due to the influence of factors such as imaging conditions and the like, the radiation intensity of the SAR images in the same region is greatly different, and when large-area SAR image stitching is performed, a better visual effect is obtained through a multi-image color homogenizing technology.
The currently common image homogenizing methods are a histogram matching method and a Wallis filtering homogenizing method. The histogram matching method is to change the shape of the histogram of the image to be leveled so that the histogram has a mean value and a variance similar to those of the reference image. The Wallis filtering method adjusts through the respective mean and variance of the images so that the mean and variance of the images to be leveled are close to the reference image.
However, the conventional color homogenizing method has the following technical problems: the histogram matching method modifies the shape of the image, and when the difference between the image to be uniformly colored and the reference image is large, the relative distance between different gray levels of the image can be changed, and color cast is easy to cause. When the Wallis filtering method is used for homogenizing the SAR image containing the water area, the image after the homogenization is also brighter or darker due to the existence of the water area, so that the color cast is caused. Because SAR aerial band images are large in coverage area and numerous in ground objects, the vision effect after the color homogenization is affected due to the unbalanced radiation intensity of the aerial band images, whether the color homogenization is carried out according to global information or the overlapping area information.
Disclosure of Invention
The invention provides an image homogenizing method and device based on an image overlapping area, wherein after an image of a navigation belt to be homogenized is obtained by the method, the overlapping area between the images and radiation characteristic curve information of the image of the navigation belt to be homogenized can be identified, and the homogenizing is carried out based on the overlapping area and the radiation characteristic curve information, so that the deviation of the homogenizing is reduced, and the homogenizing effect is improved.
A first aspect of an embodiment of the present invention provides an image blending method based on an image overlapping area, where the method includes:
acquiring a plurality of preprocessed aerial belt images;
extracting an overlapping region from adjacent aerial band images based on a preset shielding water area mask, and extracting a radiation characteristic curve of the adjacent aerial band images;
and carrying out first color homogenizing treatment on the adjacent avionics images according to the pixels of the overlapping area and the radiation characteristic curve.
In a possible implementation manner of the first aspect, the performing, according to the pixels of the overlapping area and the radiation characteristic curve, a first color-homogenizing process on the adjacent avionics band image includes:
counting first pixel information of the overlapping region, and performing fitting treatment on the radiation characteristic curve to obtain a radiation characteristic fitting curve, wherein the first pixel information comprises a first mean value and a first standard deviation of pixels in the overlapping region;
and carrying out first color homogenizing treatment on pixels of the adjacent avionics images corresponding to the overlapping area according to the first pixel information and the radiation characteristic fitting curve.
In a possible implementation manner of the first aspect, the calculation of the first mean value m is as follows:
the calculation of the first standard deviation sigma is shown in the following formula;
in a possible implementation manner of the first aspect, the calculation of the first color homogenizing process is as follows:
in the above, m x And m y The first mean value, sigma, of the pixels of the reference aerial band image X and the aerial band image Y to be leveled in the overlapping region x Sum sigma y Respectively referenced ribbon image X and ribbon to be leveledThe first standard deviation of the pixels in the overlapping region of image Y, beta, is the radiation characteristic fitting curve.
In a possible implementation manner of the first aspect, the threshold condition of the calculation of the first shading process is as follows:
wherein 0< gamma <1,0< alpha <1.
In a possible implementation manner of the first aspect, the fitting the radiation characteristic curve to obtain a radiation characteristic fitting curve includes:
when two adjacent aerial belt images have faults, performing curve restoration on a fault area corresponding to the radiation characteristic curve by utilizing preset linear interpolation, and performing polynomial fitting on the restored curves to obtain a radiation characteristic fitting curve;
or alternatively;
and when two adjacent aerial belt images are free of faults, performing polynomial fitting on the radiation characteristic curves to obtain radiation characteristic fitting curves.
In a possible implementation manner of the first aspect, the preprocessing includes:
extracting image information from a preset data set, wherein the preset data set comprises a plurality of initial images to be spliced, and the image information comprises geographic position information and track information of each initial image;
screening a plurality of initial images belonging to the same navigation band from a preset data set according to the track information to serve as co-track images;
and sequencing the plurality of same-rail images corresponding to each navigation belt according to the geographic position information, and performing second color homogenizing treatment on the plurality of same-rail images corresponding to each navigation belt according to the Euclidean distance from the center point of the navigation belt to obtain the navigation belt images.
In a possible implementation manner of the first aspect, the performing a second color-homogenizing process on the plurality of same-track images corresponding to each band according to the euclidean distance from the center point of the band to obtain a band image includes:
acquiring and counting second pixel information of a plurality of same-track images in an image overlapping region, wherein the second pixel information comprises a second average value and a second standard deviation of pixels in the image overlapping region;
and taking Euclidean distance from the center point of the navigation belt as the splicing sequence, and carrying out second color homogenizing treatment on a plurality of same-rail images corresponding to each navigation belt according to the second pixel information.
In a possible implementation manner of the first aspect, the calculation of the second mean value m is as follows:
the calculation of the second standard deviation sigma is shown in the following formula;
in the above formula, I is the co-track image, M and N are the length and width of the co-track image, respectively, (x, y) is the coordinates of the pixel in the image.
The calculation of the second color homogenizing treatment is as follows:
in the above, m x And m y The second average value, sigma, of pixels in the image overlapping region of the reference co-rail image X and the co-rail image Y to be leveled respectively x Sum sigma y The reference co-rail image X and the co-rail image Y to be leveled are respectively the second standard deviation of pixels in the image overlapping area.
A second aspect of an embodiment of the present invention provides an image blending apparatus based on an image overlapping area, the apparatus including:
the image acquisition module is used for acquiring a plurality of preprocessed navigation belt images;
the extraction module is used for extracting an overlapping area from the adjacent aerial belt images based on a preset shielding water area mask and extracting a radiation characteristic curve of the adjacent aerial belt images;
and the image homogenizing processing module is used for carrying out first homogenizing processing on the adjacent avionics images according to the pixels of the overlapping area and the radiation characteristic curve.
Compared with the prior art, the image color homogenizing method and device based on the image overlapping area provided by the embodiment of the invention have the beneficial effects that: according to the invention, a plurality of preprocessed aerial belt images can be obtained, the two aerial belt images are subjected to color homogenizing treatment according to pixel information of the overlapped area and a radiation characteristic fitting curve of the aerial belt images to be subjected to color homogenizing on the basis of a preset mask of a shielding water area in an overlapped area of the two adjacent aerial belt images, so that the deviation of color homogenizing is reduced, and the color homogenizing effect is improved.
Drawings
Fig. 1 is a flowchart of an image blending method based on an image overlapping area according to an embodiment of the present invention;
FIG. 2 is a schematic diagram of a SAR image set for multiple images provided in accordance with one embodiment of the present invention;
FIG. 3 is a schematic illustration of one of the dataset images provided in accordance with an embodiment of the present invention;
FIG. 4 is a schematic illustration of one of the preprocessed aerial belt images according to an embodiment of the present invention;
fig. 5 is a schematic view of a masking water mask for aerial images according to an embodiment of the present invention.
FIG. 6 is a schematic view showing the effect of one band image of a fault provided by an embodiment of the present invention;
FIG. 7 is a schematic representation of a smoothed radiation characteristic fit curve provided by an embodiment of the present invention;
FIG. 8 is a schematic diagram of a color-homogenizing and stitching sequence of co-rail images according to an embodiment of the present invention;
FIG. 9 is an operation flow chart of an image blending method based on an image overlapping area according to an embodiment of the present invention;
fig. 10 is a schematic structural diagram of an image blending device based on an image overlapping area according to an embodiment of the present invention.
Detailed Description
The following description of the embodiments of the present invention will be made clearly and completely with reference to the accompanying drawings, in which it is apparent that the embodiments described are only some embodiments of the present invention, but not all embodiments. All other embodiments, which can be made by those skilled in the art based on the embodiments of the invention without making any inventive effort, are intended to be within the scope of the invention.
In order to solve the above-mentioned problems, an image homogenizing method based on an image overlapping area provided in the embodiment of the present application will be described and illustrated in detail by the following specific embodiments.
Referring to fig. 1, a flowchart of an image homogenizing method based on an image overlapping area according to an embodiment of the present invention is shown.
Wherein, as an example, the image homogenizing method based on the image overlapping area may include:
s11, acquiring a plurality of preprocessed aerial belt images.
In an embodiment, a data set or an image set may be acquired, where the data set or the image set is composed of SAR images, and the multiple SAR images may be initial images belonging to the same geographic location or the same area, and the multiple initial images of the image set are preprocessed to generate multiple band images, and the multiple band images may be images belonging to the same geographic location or the same area.
Wherein, as an example, the preprocessing may comprise the sub-steps of:
s111, extracting image information from a preset data set, wherein the preset data set comprises a plurality of initial images to be spliced, and the image information comprises geographic position information and track information of each initial image.
Referring to fig. 2-3, a schematic diagram of a plurality of image set SAR images provided by an embodiment of the present invention and a schematic diagram of one of the data set images provided by an embodiment of the present invention are shown, respectively.
In an actual operation, the file name of each initial image can be read, and then the geographical position information and the track information of the SAR image contained in the file name are extracted to obtain the image information of each initial image.
S112, screening a plurality of initial images belonging to the same navigation band from a preset data set according to the track information to serve as the same track image.
S113, sorting the same-track images corresponding to each navigation belt according to the geographic position information, and carrying out second color homogenizing treatment on the same-track images corresponding to each navigation belt according to Euclidean distance from the center point of the navigation belt to obtain the navigation belt images.
In an embodiment, the file name already contains the location information and the track information, and the initial image in the same navigation band can be found according to the track information to be used as the co-track image. Meanwhile, the image data in the same navigation band can be sequenced according to the geographic position information. Then, the Euclidean distance between the image position and the center point of the navigation belt can be calculated according to the geographic information, then the splicing sequence is determined from the near to the far according to the Euclidean distance between the image position and the center point of the navigation belt, and a plurality of same-track images corresponding to the same navigation belt are subjected to color homogenizing treatment according to the splicing sequence.
Referring to fig. 4, a schematic diagram of one of the preprocessed aerial belt images is shown, according to an embodiment of the present invention.
The second color-homogenizing process may be performed on a plurality of co-rail images corresponding to the same navigation belt according to the stitching sequence, so as to form a navigation belt image, as shown in fig. 4. The ribbon images of different ribbons are combined to form a dataset of images.
In an embodiment, adjacent images in the same navigation band are overlapped, so that the images can be connected together according to the geographical position information order, and the images of the same navigation band can be subjected to color-homogenizing and splicing.
Optionally, the file name of the image also contains longitude and latitude information of the SAR image, longitude and latitude information of all images in the navigation band can be extracted to obtain the maximum value and the minimum value of the longitude and latitude of all the images, the longitude range and the latitude range of the whole area are obtained, the middle latitude is taken, the latitude difference between each image in all the navigation bands and the middle latitude is calculated, the images in the navigation bands are sequenced from small to large according to the difference value, and the splicing sequence is determined.
In one embodiment, adjacent ribbon images overlap, so that the individual ribbon images may be ordered according to geographic location information such that they are adjacent. Alternatively, longitude and latitude information of all the ribbon images can be extracted to obtain maximum values and minimum values of the longitudes and the latitudes of all the images, longitude ranges and latitude ranges of the whole area are obtained, intermediate longitudes are taken, longitude differences of all the ribbon images and the intermediate longitudes are calculated, the ribbon images are sequenced according to the difference values from small to large, and the splicing sequence is determined.
In order to enable more uniform color of images of the same swath when stitched, step S113 may include, as an example, the following sub-steps:
s1131, acquiring and counting second pixel information of a plurality of same-track images in an image overlapping region, wherein the second pixel information comprises a second average value and a second standard deviation of pixels in the overlapping region.
In an embodiment, a suitable threshold may be set to generate a mask of the shielded water, and the mask of the shielded water is used to extract the overlapping position of the two adjacent images, so as to obtain the overlapping region of the images.
In one implementation, an image overlapping area between every two of the plurality of same-track images belonging to the same navigation belt may be determined first, and then second pixel information in the area may be acquired. Wherein the second pixel information includes a second mean and a second standard deviation of pixels in the image overlapping region.
In one embodiment, the second mean value m is calculated as follows:
the calculation of the second standard deviation sigma is shown in the following formula;
in the above formula, I is the co-track image, M and N are the length and width of the co-track image, respectively, (x, y) is the coordinates of the pixel in the image.
S1132, performing second color homogenizing treatment on the plurality of same-track images corresponding to each navigation belt according to the second pixel information by taking Euclidean distance from the center point of the navigation belt as the splicing sequence.
In an implementation manner, the sequence of splicing can be determined according to the Euclidean distance from the center point of the navigation belt, and then a plurality of same-rail images corresponding to the same navigation belt are subjected to second color homogenizing processing according to the second pixel information according to the sequence of splicing, so that the navigation belt images are formed.
In one embodiment, the second color homogenizing process is calculated as follows:
in the above, m x And m y The second average value, sigma, of pixels in the image overlapping region of the reference co-rail image X and the co-rail image Y to be leveled respectively x Sum sigma y The reference co-rail image X and the co-rail image Y to be leveled are respectively the second standard deviation of pixels in the image overlapping area.
S12, extracting an overlapping area from the adjacent aerial belt images based on a preset shielding water area mask, and extracting a radiation characteristic curve of the adjacent aerial belt images.
In one embodiment, appropriate thresholds are set to generate a mask of the shielded waters, which excludes the effects of the waters in subsequent pixel information statistics.
Referring to fig. 5, a schematic diagram of a masking body mask for masking a aerial image according to an embodiment of the present invention is shown.
In this embodiment, the threshold value may be generally set to 0 to 20 according to the actual application requirement or the actual image situation. As shown in fig. 5, 15 is used as a threshold value, and pixels having a gradation equal to or higher than the threshold value in the image are white and black. When the pixel information is counted, only the pixel information of the white area is counted, and the black part in the mask is shielded to form a shielded water area mask.
By arranging the shielding water area mask, the shielding water area mask can be utilized to eliminate the water area in the image, so that the influence of the water body on the subsequent overlapping operation of the image is eliminated.
After the water areas in the images are eliminated, the overlapping areas of the two adjacent images in the order can be acquired. The spatial relationship between the reference image and the image to be leveled can be judged through the overlapping area.
In one implementation, the geographic location information attached to the image may be obtained, and if the latitude and longitude ranges of the two images overlap, an overlapping area is considered to exist, and the area may be extracted according to the geographic location information to obtain the overlapping area.
Optionally, if two adjacent images after sequencing have no coincident pixels, the two images can be used to carry out subsequent information statistics by using full image pixels; if one image contains the other image, carrying out subsequent information statistics on the whole pixels of the contained image, and using the pixels of the region where the contained image is located as an overlapping region for the other image; and if the two images are adjacent, determining an overlapping region of the two images according to the geographic position information attached to the SAR image.
In general, the area of the real overlapping area is relatively small compared with the whole image, in order to obtain more image information, the periphery of the real overlapping area of each image is expanded to a certain extent, and the overlapping area is added with the expanding area to be used as the overlapping area in the algorithm.
The overlapping region in this step is extracted in the same manner as the image overlapping region in the above step.
In an embodiment, the radiation characteristic curve may be a radiation characteristic curve for counting the flight direction of the aerial image on the imaging platform, and the influence of the water body in the subsequent color homogenizing treatment can be eliminated through the radiation characteristic, so that the color homogenizing effect is further improved.
S13, performing first color homogenizing treatment on the adjacent avionics images according to the pixels of the overlapping area and the radiation characteristic curve.
In an embodiment, pixel information of the overlapping area can be obtained, and then two images of the aviation belt corresponding to the overlapping area are subjected to color homogenizing treatment according to the pixel information and the radiation characteristic curve, so that the color of the overlapping area of the two images is uniform, and the situation of chromatic aberration or color cast is avoided.
In one embodiment, the plurality of sequenced avionics images may be continuous. In order to be able to rapidly homogenize a plurality of successive aerial belt images, step S13 may comprise the following sub-steps, as an example:
s131, counting first pixel information of the overlapping area, and carrying out fitting processing on the radiation characteristic curve to obtain a radiation characteristic fitting curve, wherein the first pixel information comprises a first mean value and a first standard deviation of pixels in the overlapping area.
In one embodiment, the first mean value m is calculated as follows:
the calculation of the first standard deviation sigma is shown in the following formula;
in an alternative embodiment, the first mean value and the third mean value are calculated in the same manner, and the first standard deviation and the third standard deviation are calculated in the same manner.
The radiation characteristic curve RC is represented by the following formula:
RC＝ml/mlm；
wherein,
in the above formula, I is a band image, M and N are the length and width of the band image, respectively, (x, y) are coordinates of pixels in the band image.
Referring to fig. 6, an effect diagram of one band image of a fault provided by an embodiment of the present invention is shown.
In another embodiment, the band images may or may not be connected, as shown in fig. 6, the SAR images within the unified band in the preprocessing may not be completely continuous, and one or two images may be missing in the middle, so that a fault appears in the much Zhang Hang band images. If the aerial band image with the faults is used as an image to be leveled, when the radiation characteristic curve is extracted and then is subjected to subsequent fitting to generate a radiation characteristic fitting curve, curve fitting errors near a fault area can be increased due to the faults. In order to better perform the color-homogenizing process on the tomographic image, the step S131 may include the following sub-steps, as an example:
s1311, when two adjacent aerial belt images have faults, performing curve restoration on a fault area corresponding to the radiation characteristic curve by utilizing preset linear interpolation, and performing polynomial fitting on the restored curve to obtain a radiation characteristic fitting curve;
or alternatively;
s1312, when two adjacent aerial belt images are free of faults, performing polynomial fitting on the radiation characteristic curves to obtain radiation characteristic fitting curves.
In a specific operation, if two adjacent aerial belt images are free of faults, polynomial fitting can be directly carried out on the radiation characteristic curve to obtain a radiation characteristic fitting curve. And then carrying out subsequent color homogenizing operation by utilizing the radiation characteristic fitting curve.
If two adjacent aerial belt images have faults, curve restoration can be carried out on a fault area corresponding to the radiation characteristic curve by utilizing preset linear interpolation, and a restored radiation characteristic curve is obtained. And then polynomial fitting is carried out on the repaired curve to obtain a radiation characteristic fitting curve.
Referring to fig. 7, a schematic diagram of a smooth radiation characteristic fitting curve provided by an embodiment of the present invention is shown.
Due to the problem of missing image data, etc., the image data within the same swath is not necessarily continuous. Thus, the radiation characteristic fitted curve can be detected, and if a "fault" exists, linear interpolation is used to perform curve repair on the fault region of the radiation characteristic fitted curve.
It should be noted that, the missing image cannot be restored, but the problem caused by suddenly decreasing the radiation characteristic fitting curve to zero due to the missing image can be solved.
The radiation intensity in the air belt is not uniform due to the reasons of photographing time, photographing angle and the like, polynomial fitting is used for the repaired radiation characteristic fitting curve, the higher the polynomial degree is, the stronger the self-adaptability of the color homogenizing scheme is, but the higher the polynomial degree is, the relative gray level between special features (such as mountains, villages, cities and the like) and other features can be greatly reduced, so that the proper number n of terms is selected in the experimental process, and a smooth radiation characteristic fitting curve beta is obtained, as shown in fig. 7.
Polynomial fitting refers to fitting a polynomial p (x) =p 1 x n +p 2 x n-1 +…+p n x+p n+1 Fitting a radiation characteristic fitting curve, which is intended to extract a smoothed radiation characteristic "trend" from the non-smoothed radiation characteristic fitting curve, resulting in a smoothed radiation characteristic fitting curve β.
And S132, performing first color homogenizing treatment on pixels of the adjacent avionics images corresponding to the overlapping area according to the first pixel information and the radiation characteristic fitting curve.
The calculation of the first color homogenizing treatment is as follows:
in the above, m x And m y The first mean value, sigma, of the pixels of the reference aerial band image X and the aerial band image Y to be leveled in the overlapping region x Sum sigma y And respectively referring to the first standard deviation of pixels in the overlapping area of the avionic image X and the avionic image Y to be leveled, wherein beta is a smooth radiation characteristic fitting curve.
It should be noted that, during the color homogenizing, although the influence of the water body is eliminated by the mask of the preset mask water area mask in the previous step, in order to ensure that the consistency of the histogram and the relative distance between different gray levels do not change too much during the color homogenizing process, the color homogenizing process is also performed on the water area, so that the color uniformity of each image can be ensured.
By means of image evening processing, the colors of two (or more) ribbon images can be made to be consistent, and for two-to-two ribbon images to be even, the color of one image (the ribbon image to be even) is required to be close to that of the other image (the reference ribbon image). The reference image may be a first band image, and the image to be leveled may be a second adjacent band image.
In the process of the evening, the aerial belt images to be evening are not only overlapped, but the whole aerial belt images need to be subjected to the evening.
The method can iterate for a certain number of times when the image is uniformly colored, so that the color-uniform effect is better.
The reference image is related to the splicing sequence, the image closest to the center point of the latitude of the navigation belt is selected as an initial reference image by first splicing in the navigation belt during preprocessing, and the second closest image is used as an image to be leveled; the nth stitching selects the (N-1) th color-homogenizing stitching result image as a reference image, and the image closest to the (N) th center point is used as an image to be color-homogenized, as shown in fig. 8.
In one embodiment, the imaging platform flight direction may be set along the x-axis direction.
It can be seen that by adding the influence of the smooth radiation characteristic fitting curve beta, the radiation intensity is relatively strong, and the pixel value is inhibited after the homogenization in a relatively bright area, so that the whole image of the navigation belt is more harmonious in visual sense after the homogenization.
Because the coverage area of the image of the navigation belt is usually larger, in order to avoid that special areas (extremely bright areas and extremely dark areas) in the navigation belt go to the other extreme after being uniformly colored, a threshold is set to limit the color-uniformly effect.
Optionally, in this embodiment, the threshold condition of the calculation of the first color matching process is as follows, whether the image has a fault or no fault:
wherein, 0< gamma <1,0< alpha <1, can be adjusted according to the actual situation. Preferably, γ=0.3, α=0.8.
In the leveling process, the image to be leveled is not just an overlapping area, but an entire image to be leveled. Reference is made in particular to the description above.
In addition, when the formula is adopted for the color homogenizing treatment, in order to ensure that the consistency of the histogram and the relative distance between different gray levels are not changed greatly, the color homogenizing treatment is also carried out on a water area.
In an embodiment, the above operation method may be used to perform color-homogenizing and stitching on multiple aerial belt images to form a large-area SAR image
Specifically, the uniformly colored ribbon images can be spliced in pairs according to the geographic position information attached to each SAR ribbon image until the image splicing of the whole large area is completed.
Referring to fig. 9, an operation flowchart of an image blending method based on an image overlapping area according to an embodiment of the present invention is shown.
Specifically, data (including a plurality of SAR images) may be read in; preprocessing a plurality of SAR initial images (specifically, preprocessing can be to acquire an image overlapping area of two adjacent SAR initial images belonging to the same navigation belt and counting pixel information of the area, and carrying out image color homogenization and image stitching on the two adjacent images according to the pixel information) to obtain a plurality of navigation belt images; then preprocessing the multiple navigation belt images; then, acquiring an overlapping area between adjacent aerial belt images, and counting pixel information in the overlapping area; counting a radiation characteristic fitting curve of the aerial belt image to be uniformly colored; carrying out uniform color treatment on the image by utilizing a radiation characteristic fitting curve; and finally, splicing the plurality of uniformly-colored aerial belt images to obtain a result image.
Compared with the prior art, the method can acquire the overlapping area of the adjacent images, and the ground features of the overlapping area of the adjacent images are similar, so that the color homogenizing method based on the overlapping area is more reasonable, and the color homogenizing effect is better; the elimination of the water area information reduces the color uniformity interference caused by the water body; in addition, when splicing among the navigation belts, the color homogenizing algorithm dynamically adjusts according to the radiation intensities of different areas on the same navigation belt, so that the whole color homogenizing image is more harmonious in vision.
In this embodiment, the embodiment of the present invention provides an image homogenizing method based on an image overlapping area, which has the following beneficial effects: according to the invention, a plurality of preprocessed aerial belt images can be obtained, an overlapping area is extracted from two adjacent aerial belt images based on a preset mask of a shielding water area, and the aerial belt images to be leveled are leveled according to pixel information of the overlapping area and radiation characteristics of the aerial belt images to be leveled, so that the color leveling deviation is reduced, and the color leveling effect is improved.
The embodiment of the invention also provides an image homogenizing device based on the image overlapping area, and referring to fig. 10, a schematic structural diagram of the image homogenizing device based on the image overlapping area is shown.
Wherein, as an example, the image homogenizing device based on the image overlapping area may include:
an image acquisition module 101, configured to acquire a plurality of preprocessed aerial belt images;
the extracting module 102 is configured to extract an overlapping region in the adjacent aerial belt images based on a preset shielding water area mask, and extract a radiation characteristic curve of the adjacent aerial belt images;
and the image homogenizing processing module 103 is used for performing first homogenizing processing on the adjacent avionics images according to the pixels of the overlapping area and the radiation characteristic curve.
Optionally, the image shading processing module is further configured to:
counting first pixel information of the overlapping region, and performing fitting treatment on the radiation characteristic curve to obtain a radiation characteristic fitting curve, wherein the first pixel information comprises a first mean value and a first standard deviation of pixels in the overlapping region;
and carrying out first color homogenizing treatment on pixels of the adjacent avionics images corresponding to the overlapping area according to the first pixel information and the radiation characteristic fitting curve.
Optionally, the calculation of the first mean value m is as follows:
the calculation of the first standard deviation sigma is shown in the following formula;
in a possible implementation manner of the first aspect, the calculation of the first color homogenizing process is as follows:
in the above, m x And m y The first average of pixels in the overlapping region of the reference aerial band image X and the aerial band image Y to be leveled respectively,σ x sum sigma y The respectively referenced aerial band image X and the aerial band image Y to be leveled have a first standard deviation of pixels in the overlapping area, and beta is a smooth radiation characteristic fitting curve.
Optionally, the calculated threshold condition of the first shading process is as follows:
wherein 0< gamma <1,0< alpha <1.
Optionally, the image shading processing module is further configured to:
when two adjacent aerial belt images have faults, performing curve restoration on a fault area corresponding to the radiation characteristic curve by utilizing preset linear interpolation, and performing polynomial fitting on the restored curves to obtain a radiation characteristic fitting curve;
or alternatively;
and when two adjacent aerial belt images are free of faults, performing polynomial fitting on the radiation characteristic curves to obtain radiation characteristic fitting curves.
Optionally, the image acquisition module is further configured to:
extracting image information from a preset data set, wherein the preset data set comprises a plurality of initial images to be spliced, and the image information comprises geographic position information and track information of each initial image;
screening a plurality of initial images belonging to the same navigation band from a preset data set according to the track information to serve as co-track images;
and sequencing the plurality of same-rail images corresponding to each navigation belt according to the geographic position information, and performing second color homogenizing treatment on the plurality of same-rail images corresponding to each navigation belt according to the Euclidean distance from the center point of the navigation belt to obtain the navigation belt images.
Optionally, the image acquisition module is further configured to:
acquiring and counting second pixel information of a plurality of same-track images in an image overlapping region, wherein the second pixel information comprises a second average value and a second standard deviation of pixels in the image overlapping region;
and taking Euclidean distance from the center point of the navigation belt as the splicing sequence, and carrying out second color homogenizing treatment on a plurality of same-rail images corresponding to each navigation belt according to the second pixel information.
Optionally, the calculation of the second mean value m is as follows:
the calculation of the second standard deviation sigma is shown in the following formula;
in the above formula, I is the co-track image, M and N are the length and width of the co-track image, respectively, (x, y) is the coordinates of the pixel in the image.
The calculation of the second color homogenizing treatment is as follows:
in the above, m x And m y The second average value, sigma, of pixels in the image overlapping region of the reference co-rail image X and the co-rail image Y to be leveled respectively x Sum sigma y The reference co-rail image X and the co-rail image Y to be leveled are respectively the second standard deviation of pixels in the image overlapping area.
It will be clearly understood by those skilled in the art that, for convenience and brevity, the specific working process of the apparatus described above may refer to the corresponding process in the foregoing method embodiment, which is not described herein again.
Further, an embodiment of the present application further provides an electronic device, including: a memory, a processor and a computer program stored on the memory and executable on the processor, which when executed implements the image blending method based on the image overlapping area as described in the above embodiments.
Further, the embodiment of the application also provides a computer-readable storage medium storing a computer-executable program for causing a computer to execute the image blurring method based on the image overlapping region according to the above embodiment.
While the foregoing is directed to the preferred embodiments of the present invention, it will be appreciated by those skilled in the art that changes and modifications may be made without departing from the principles of the invention, such changes and modifications are also intended to be within the scope of the invention.
Claims (8)
1. An image blending method based on an image overlapping area, the method comprising:
acquiring a plurality of preprocessed aerial belt images;
extracting an overlapping region from adjacent aerial band images based on a preset shielding water area mask, and extracting a radiation characteristic curve of the adjacent aerial band images;
performing first color homogenizing treatment on the adjacent aerial belt images according to the pixels of the overlapping area and the radiation characteristic curve;
the calculation of the first color homogenizing treatment is as follows:
in the above, m x And m y The first mean value, sigma, of the pixels of the reference aerial band image X and the aerial band image Y to be leveled in the overlapping region x Sum sigma y Respectively referencing the first standard deviation of pixels of the aerial band image X and the aerial band image Y to be leveled in the overlapping area, wherein beta is a radiation characteristic fitting curve;
the calculated threshold condition of the first shading process is as follows:
wherein 0< gamma <1,0< alpha <1.
2. The image blending method based on the image overlapping area as claimed in claim 1, wherein said performing a first blending process on the adjacent avionics images according to the pixels of the overlapping area and the radiation characteristic curve comprises:
counting first pixel information of the overlapping region, and performing fitting treatment on the radiation characteristic curve to obtain a radiation characteristic fitting curve, wherein the first pixel information comprises a first mean value and a first standard deviation of pixels in the overlapping region;
and carrying out first color homogenizing treatment on pixels of the adjacent avionics images corresponding to the overlapping area according to the first pixel information and the radiation characteristic fitting curve.
3. The image blending method based on the image overlapping area as claimed in claim 2, wherein the calculation of the first mean value m is as follows:
the calculation of the first standard deviation sigma is shown in the following formula;
4. the image homogenizing method based on the image overlapping region according to any one of claims 2 to 3, wherein the fitting the radiation characteristic curve to obtain a radiation characteristic fitting curve comprises:
when two adjacent aerial belt images have faults, performing curve restoration on a fault area corresponding to the radiation characteristic curve by utilizing preset linear interpolation, and performing polynomial fitting on the restored curves to obtain a radiation characteristic fitting curve;
or alternatively;
and when two adjacent aerial belt images are free of faults, performing polynomial fitting on the radiation characteristic curves to obtain radiation characteristic fitting curves.
5. A method of image blending based on image overlapping areas according to any of claims 1-3, wherein said preprocessing comprises:
extracting image information from a preset data set, wherein the preset data set comprises a plurality of initial images to be spliced, and the image information comprises geographic position information and track information of each initial image;
screening a plurality of initial images belonging to the same navigation band from a preset data set according to the track information to serve as co-track images;
and sequencing the plurality of same-rail images corresponding to each navigation belt according to the geographic position information, and performing second color homogenizing treatment on the plurality of same-rail images corresponding to each navigation belt according to the Euclidean distance from the center point of the navigation belt to obtain the navigation belt images.
6. The image homogenizing method based on the image overlapping area of claim 5, wherein the performing the second homogenizing process on the plurality of co-rail images corresponding to each band according to the euclidean distance from the center point of the band to obtain the band image comprises:
acquiring and counting second pixel information of a plurality of same-track images in an image overlapping region, wherein the second pixel information comprises a second average value and a second standard deviation of pixels in the image overlapping region;
and taking Euclidean distance from the center point of the navigation belt as the splicing sequence, and carrying out second color homogenizing treatment on a plurality of same-rail images corresponding to each navigation belt according to the second pixel information.
7. The image blending method based on the image overlapping area as claimed in claim 6, wherein the calculation of the second mean value m is as follows:
the calculation of the second standard deviation sigma is shown in the following formula;
in the above formula, I is the same-track image, M and N are the length and the width of the same-track image respectively, and (x, y) is the coordinates of pixels in the image;
the calculation of the second color homogenizing treatment is as follows:
in the above, m x And m y The second average value, sigma, of pixels in the image overlapping region of the reference co-rail image X and the co-rail image Y to be leveled respectively x Sum sigma y The reference co-rail image X and the co-rail image Y to be leveled are respectively the second standard deviation of pixels in the image overlapping area.
8. An image blending apparatus based on an image overlapping region, the apparatus comprising:
the image acquisition module is used for acquiring a plurality of preprocessed navigation belt images;
the extraction module is used for extracting an overlapping area from the adjacent aerial belt images based on a preset shielding water area mask and extracting a radiation characteristic curve of the adjacent aerial belt images;
the image homogenizing processing module is used for performing first homogenizing processing on the adjacent aerial belt images according to the pixels of the overlapping area and the radiation characteristic curve;
the calculation of the first color homogenizing treatment is as follows:
in the above, m x And m y The first mean value, sigma, of the pixels of the reference aerial band image X and the aerial band image Y to be leveled in the overlapping region x Sum sigma y Respectively referencing the first standard deviation of pixels of the aerial band image X and the aerial band image Y to be leveled in the overlapping area, wherein beta is a radiation characteristic fitting curve;
the calculated threshold condition of the first shading process is as follows:
wherein 0< gamma <1,0< alpha <1.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310372537.XA CN116503274B (en) | 2023-04-07 | 2023-04-07 | Image color homogenizing method and device based on image overlapping area |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202310372537.XA CN116503274B (en) | 2023-04-07 | 2023-04-07 | Image color homogenizing method and device based on image overlapping area |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN116503274A CN116503274A (en) | 2023-07-28 |
| CN116503274B true CN116503274B (en) | 2023-12-22 |
Family
ID=87323923
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202310372537.XA Active CN116503274B (en) | 2023-04-07 | 2023-04-07 | Image color homogenizing method and device based on image overlapping area |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN116503274B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN117541469A (en) * | 2024-01-10 | 2024-02-09 | 中山大学 | SAR image stitching method and device based on graph theory |
Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107230186A (en) * | 2016-03-23 | 2017-10-03 | 中国科学院遥感与数字地球研究所 | A kind of even color method of satellite remote-sensing image physics |
| CN107480727A (en) * | 2017-08-28 | 2017-12-15 | 荆门程远电子科技有限公司 | The unmanned plane image fast matching method that a kind of SIFT and ORB are combined |
| CN108564532A (en) * | 2018-03-30 | 2018-09-21 | 合肥工业大学 | Large scale distance satellite-borne SAR image method for embedding |
| CN108804394A (en) * | 2018-05-23 | 2018-11-13 | 河海大学 | A kind of construction method of city noctilucence total amount-urban population regression model |
| CN108875583A (en) * | 2018-05-23 | 2018-11-23 | 河海大学 | A kind of TNL-GDP regression model construction method based on urban architecture classification |
| CN109685081A (en) * | 2018-12-27 | 2019-04-26 | 中国土地勘测规划院 | A kind of joint change detecting method of Remotely sensed acquisition black fallow |
| CN111091502A (en) * | 2019-11-11 | 2020-05-01 | 苏州中科天启遥感科技有限公司 | Remote sensing image color homogenizing method and system, storage medium and electronic equipment |
| CN111652826A (en) * | 2020-05-18 | 2020-09-11 | 哈尔滨工业大学 | Multiple multi/hyperspectral remote sensing image color homogenizing method based on Wallis filtering and histogram matching |
| CN111680704A (en) * | 2020-06-11 | 2020-09-18 | 生态环境部卫星环境应用中心 | Automatic and rapid extraction method and device for newly-increased human active plaque of ocean red line |
| CN111709901A (en) * | 2020-05-22 | 2020-09-25 | 哈尔滨工业大学 | Non-multiple multi/hyperspectral remote sensing image color homogenizing method based on FCM cluster matching and Wallis filtering |
| CN112765095A (en) * | 2020-12-24 | 2021-05-07 | 山东省国土测绘院 | Method and system for filing image data of stereo mapping satellite |
| CN114998399A (en) * | 2022-05-20 | 2022-09-02 | 中国人民解放军61540部队 | Heterogeneous optical remote sensing satellite image stereopair preprocessing method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US7613335B2 (en) * | 2003-02-12 | 2009-11-03 | The University Of Iowa Research Foundation | Methods and devices useful for analyzing color medical images |
| US10055850B2 (en) * | 2014-09-19 | 2018-08-21 | Brain Corporation | Salient features tracking apparatus and methods using visual initialization |
-
2023
- 2023-04-07 CN CN202310372537.XA patent/CN116503274B/en active Active
Patent Citations (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN107230186A (en) * | 2016-03-23 | 2017-10-03 | 中国科学院遥感与数字地球研究所 | A kind of even color method of satellite remote-sensing image physics |
| CN107480727A (en) * | 2017-08-28 | 2017-12-15 | 荆门程远电子科技有限公司 | The unmanned plane image fast matching method that a kind of SIFT and ORB are combined |
| CN108564532A (en) * | 2018-03-30 | 2018-09-21 | 合肥工业大学 | Large scale distance satellite-borne SAR image method for embedding |
| CN108804394A (en) * | 2018-05-23 | 2018-11-13 | 河海大学 | A kind of construction method of city noctilucence total amount-urban population regression model |
| CN108875583A (en) * | 2018-05-23 | 2018-11-23 | 河海大学 | A kind of TNL-GDP regression model construction method based on urban architecture classification |
| CN109685081A (en) * | 2018-12-27 | 2019-04-26 | 中国土地勘测规划院 | A kind of joint change detecting method of Remotely sensed acquisition black fallow |
| CN111091502A (en) * | 2019-11-11 | 2020-05-01 | 苏州中科天启遥感科技有限公司 | Remote sensing image color homogenizing method and system, storage medium and electronic equipment |
| CN111652826A (en) * | 2020-05-18 | 2020-09-11 | 哈尔滨工业大学 | Multiple multi/hyperspectral remote sensing image color homogenizing method based on Wallis filtering and histogram matching |
| CN111709901A (en) * | 2020-05-22 | 2020-09-25 | 哈尔滨工业大学 | Non-multiple multi/hyperspectral remote sensing image color homogenizing method based on FCM cluster matching and Wallis filtering |
| CN111680704A (en) * | 2020-06-11 | 2020-09-18 | 生态环境部卫星环境应用中心 | Automatic and rapid extraction method and device for newly-increased human active plaque of ocean red line |
| CN112765095A (en) * | 2020-12-24 | 2021-05-07 | 山东省国土测绘院 | Method and system for filing image data of stereo mapping satellite |
| CN114998399A (en) * | 2022-05-20 | 2022-09-02 | 中国人民解放军61540部队 | Heterogeneous optical remote sensing satellite image stereopair preprocessing method |
Non-Patent Citations (1)
| Title |
|---|
| 基于改进Wallis滤波器算法的无人机影像匀光匀色方法;杨海成等;《北京测绘》;全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN116503274A (en) | 2023-07-28 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8180173B2 (en) | Flash artifact eye defect correction in blurred images using anisotropic blurring | |
| US7095420B2 (en) | System and related methods for synthesizing color imagery | |
| US8457437B2 (en) | System and method for enhancing registered images using edge overlays | |
| US6898331B2 (en) | Image fusion system and method | |
| CN116503274B (en) | Image color homogenizing method and device based on image overlapping area | |
| CN110827218B (en) | Airborne image defogging method based on weighted correction of HSV (hue, saturation, value) transmissivity of image | |
| Soderblom et al. | Global color variations on the Martian surface | |
| CN113343789A (en) | High-resolution remote sensing image land cover classification method based on local detail enhancement and edge constraint | |
| WO2010022029A1 (en) | Method and apparatus for filling in or replacing image pixel data | |
| Kauth et al. | Blob: An unsupervised clustering approach to spatial preprocessing of mss imagery | |
| CN112233190B (en) | Satellite remote sensing image color balancing method based on block adjustment | |
| CN112686172B (en) | Airport runway foreign matter detection method, device and storage medium | |
| Lisani et al. | An inquiry on contrast enhancement methods for satellite images | |
| CN112101309A (en) | Ground object target identification method and device based on deep learning segmentation network | |
| CN112307901A (en) | Landslide detection-oriented SAR and optical image fusion method and system | |
| CN114202491B (en) | Method and system for enhancing optical image | |
| Wang et al. | Settlement extraction in the North China Plain using Landsat and Beijing-1 multispectral data with an improved watershed segmentation algorithm | |
| Basso et al. | Combining morphological filtering, anisotropic diffusion and block-based data replication for automatically detecting and recovering unscanned gaps in remote sensing images | |
| CN107507151B (en) | Multispectral remote sensing image real color restoration method and system | |
| CN112861774A (en) | Method and system for identifying ship target by using remote sensing image | |
| Widyaningsih et al. | Optimization Contrast Enhancement and Noise Reduction for Semantic Segmentation of Oil Palm Aerial Imagery. | |
| Lewiński et al. | Testing texture of VHR panchromatic data as a feature of land cover classification | |
| Ichikawa et al. | Identification of paddy fields in Northern Japan using RapidEye images | |
| Fassett et al. | Age, Evolution, and Dispersion of the Loose Groups of Blue Stars in the Northeast Radio Lobe of Centaurus A | |
| CN108053406B (en) | Surface coverage mapping method and device based on multi-resolution remote sensing image |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |