A method and system for fast image fusion based on visual field information

Abstract

The application discloses a kind of based on visual field information's image fast fusion method and system, this method includes: obtaining the boundary line of multiple registered to-be-fused images, calculate the shortest distance between each overlapping pixel point in the overlapping area of multiple the to-be-fused image and corresponding to-be-fused image boundary line respectively, determine the contribution weight of each to-be-fused image to each overlapping pixel point, according to the contribution weight of each to-be-fused image to each overlapping pixel point and the pixel value of each overlapping pixel point in corresponding to-be-fused image respectively, determine the fusion pixel of each overlapping pixel point, obtain fusion image.The spatial position of pixel is introduced in the application, so that the contribution weight of to-be-fused image to overlapping pixel is more accurate, the fused pixel value is calculated by the contribution weight of overlapping pixel and the pixel value of overlapping pixel in different to-be-fused image respectively, improve image fusion efficiency and effect, reduce the generation of motion artifact.

Description

Technical Field

[0001] This invention relates to the field of image fusion technology, and specifically to a method and system for rapid image fusion based on visual field information. Background Technology

[0002] Traditional image fusion methods mainly include image fusion algorithms based on extreme values, image fusion algorithms based on pixel-weighted averages, and image fusion algorithms based on multiple scales.

[0003] Image fusion algorithms based on extreme values ​​compare the gray values ​​of corresponding pixels in the overlapping regions of two images and use the pixel with the larger (or smaller) gray value as the pixel value of the fused image. This fusion algorithm fails when the two images have large differences in brightness and darkness, and the result is close to that of no fusion, so its applicability is very limited.

[0004] Image fusion algorithms based on pixel weighted averaging calculate the pixel value of the final fused image by weighting the corresponding pixels of two images. This method only considers the grayscale value of the pixels and ignores the position of the pixels and other factors. Therefore, the fused image generated cannot retain the original image details well, loses useful information, and increases redundant information, resulting in poor visual effect of the fused image, image blurring, and artifacts of moving targets.

[0005] Multi-scale image fusion algorithms extract and analyze image features at different scales using image pyramids. Since the pyramid-based transformation process is an up / down process, all image pyramid transformations decompose the redundancy of the image. That is, in the decomposition result, the data between adjacent scales are correlated and redundant, which can lead to block artifacts in the fusion result and affect the visual effect of the fused image.

[0006] Therefore, current image fusion algorithms cannot simultaneously meet the requirements of fusing images with large differences in brightness, ensuring the accuracy of the fused image information, and guaranteeing that the fused image does not contain artifacts. Summary of the Invention

[0007] The purpose of this invention is to provide a fast image fusion method and system based on field of view information, which solves the problem of poor image fusion effect in existing methods.

[0008] The solution of the present invention to the above-mentioned technical problems is as follows:

[0009] A fast image fusion method based on field-of-view information, characterized by the following steps:

[0010] Acquire multiple registered images to be fused;

[0011] Determine the boundary lines of the multiple images to be fused;

[0012] The shortest distance between each overlapping pixel in the overlapping region of the multiple images to be fused and the corresponding boundary line of the image to be fused is obtained respectively.

[0013] The contribution weight of each image to be fused to each overlapping pixel is determined based on the shortest distance.

[0014] Obtain the pixel value of each overlapping pixel in the corresponding image to be fused. Based on the contribution weight of each image to be fused to each overlapping pixel and the pixel value of each overlapping pixel in the corresponding image to be fused, determine the fusion pixel of each overlapping pixel to obtain the fused image.

[0015] Further specifying, the step of determining the boundary lines of the multiple images to be fused specifically includes the following steps:

[0016] Perform grayscale processing on the images to be merged to obtain a grayscale image to be merged;

[0017] Add 1 to the gray value of each pixel in the grayscale image to be fused, and extend the boundary of the grayscale image to be fused by 1 pixel with a gray value of 0 to obtain the extended image to be fused.

[0018] Line detection is performed on the expanded image to be fused to obtain the boundary lines of the expanded image to be fused;

[0019] Subtract 1 from the coordinates of the endpoints of the boundary line of the expanded image to obtain the boundary line of the image to be fused.

[0020] Further specifying, the step of obtaining the shortest distance between each overlapping pixel in the overlapping region of the multiple images to be fused and the corresponding boundary line of the image to be fused specifically includes the following steps:

[0021] Select any overlapping pixel in the overlapping region of the images to be fused;

[0022] Calculate the shortest distance between the overlapping pixel and the boundary line of the image to be fused;

[0023] Repeat the above steps until the shortest distance between each overlapping pixel in the overlapping region of the image to be fused and the boundary line of the image to be fused is obtained.

[0024] Further specifying, the step of determining the contribution weight of each image to be fused to each overlapping pixel based on the shortest distance specifically includes the following steps:

[0025] The obtained shortest distance is normalized to obtain the initial contribution weight of the image to be fused to each overlapping pixel.

[0026] A cosine transform is performed on each initial contribution weight to obtain the contribution weight of the image to be fused to each overlapping pixel.

[0027] Further specifying, the step of obtaining the pixel value of each overlapping pixel in the corresponding image to be fused, determining the fusion pixel of each overlapping pixel based on the contribution weight of each image to be fused to each overlapping pixel and the pixel value of each overlapping pixel in the corresponding image to be fused, and obtaining the fused image specifically includes the following steps:

[0028] Obtain the pixel value of any overlapping pixel in the overlapping region of the image to be fused;

[0029] The fused pixel value of the overlapping pixel is obtained by summing the products of the pixel value of the overlapping pixel in the image to be fused and the contribution weight of the image to be fused to the overlapping pixel.

[0030] Repeat the above steps until the fused pixel values ​​of all overlapping pixels in the overlapping area of ​​the images to be fused are obtained, thus completing the fusion of the images to be fused.

[0031] A fast image fusion system based on field-of-view information, characterized in that it includes:

[0032] The image acquisition unit is used to acquire multiple registered images to be merged.

[0033] The image boundary determination unit is used to determine the boundary lines of multiple images to be merged respectively;

[0034] The shortest distance calculation unit between overlapping pixels and edge lines is used to obtain the shortest distance between each overlapping pixel in the overlapping region of multiple images to be fused and the corresponding boundary line of the image to be fused.

[0035] The overlapping pixel contribution weight calculation unit is used to determine the contribution weight of each image to be fused to each overlapping pixel based on the shortest distance.

[0036] The image fusion unit is used to obtain the pixel value of each overlapping pixel in the corresponding image to be fused, and determine the fusion pixel of each overlapping pixel based on the contribution weight of each image to be fused to each overlapping pixel and the pixel value of each overlapping pixel in the corresponding image to be fused, so as to obtain the fused image.

[0037] Further specifying, the edge determination unit of the image to be fused includes:

[0038] The grayscale image to be merged module is used to perform grayscale processing on the image to be merged to obtain the grayscale image to be merged.

[0039] The extended image acquisition module is used to add 1 to the gray value of each pixel in the grayscale image to be merged, and to extend the boundary of the grayscale image to be merged by 1 pixel with a gray value of 0 to obtain the extended image to be merged.

[0040] The module for obtaining the boundary line of the expanded image to be fused is used to perform line detection on the expanded image to be fused to obtain the boundary line of the expanded image to be fused.

[0041] The module for obtaining the boundary line of the image to be fused is used to subtract 1 from the coordinate values ​​of the endpoints of the extended boundary line of the image to be fused to obtain the boundary line of the image to be fused.

[0042] Further specifying, the unit for calculating the shortest distance between overlapping pixels and the edge line includes:

[0043] The module for calculating the distance between overlapping pixels and the image to be fused is used to calculate the shortest distance between each overlapping pixel and the boundary line of the image to be fused.

[0044] Further specifying, the overlapping pixel contribution weight calculation unit includes:

[0045] The initial contribution weight calculation module is used to normalize the shortest distance corresponding to each overlapping pixel to obtain the initial contribution weight of the image to be fused to each overlapping pixel.

[0046] The contribution weight calculation module is used to perform a cosine transform on each initial contribution weight to obtain the cosine contribution weight of the image to be fused to each overlapping pixel.

[0047] Further specifying, the image fusion unit includes:

[0048] The overlapping pixel acquisition module is used to acquire the pixel values ​​of overlapping pixels in the overlapping region of the image to be fused.

[0049] The overlapping pixel fusion module is used to obtain the fusion pixel value of each overlapping pixel in the overlapping region of the image to be fused by summing the product of the pixel value of each overlapping pixel in the overlapping region of the image to be fused and the contribution weight of the image to be fused to that overlapping pixel in the image to be fused, thereby completing the fusion of the images to be fused.

[0050] The beneficial effects of this invention are as follows:

[0051] 1. This invention determines the range of pixels to be fused by obtaining the boundary line of each pixel to be fused, obtains the contribution weight of the image to each overlapping pixel by the shortest distance from the overlapping similarity to the boundary line of the corresponding image to be fused, and introduces the coordinate position of the pixel, thereby making the contribution weight of the image to the overlapping pixel more accurate. Thus, the fused pixel value is calculated by the contribution weight of the overlapping pixel and the pixel value of the overlapping pixel in different images to be fused, thereby completing the image fusion, improving the image fusion efficiency, reducing the generation of motion artifacts, and improving the image fusion effect.

[0052] 2. This invention can reduce the amount and difficulty of calculation by shrinking the image to be fused, thereby improving the efficiency of image fusion. At the same time, by scaling up and down the cosine contribution weight of the image to be fused to each overlapping pixel, the contribution weight of the image to be fused to each pixel in the overlapping area can be obtained. This allows for the rapid and accurate calculation of the fused pixels in the overlapping area of ​​the two images to be fused, so that the fused image meets the usage requirements. Attached Figure Description

[0053] Figure 1 This is a flowchart illustrating the steps of the fast image fusion method based on field of view information in Embodiment 1 of the present invention.

[0054] Figure 2 Image A to be fused in this invention;

[0055] Figure 3 Image B to be fused in this invention;

[0056] Figure 4 The image obtained in Embodiment 2 of the present invention is a fused image.

[0057] Figure 5 This refers to the image obtained by fusing image A and image B using existing methods.

[0058] Figure 6 This is a schematic diagram of the image fast fusion system based on field of view information according to the present invention;

[0059] Figure 7 This is a schematic diagram of the image fast fusion system based on field of view information according to the present invention. Detailed Implementation

[0060] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0061] Example 1

[0062] refer to Figure 1 This embodiment provides a fast image fusion method based on field of view information, including the following steps:

[0063] Acquire multiple registered images to be fused;

[0064] To further explain, in actual image fusion, a large number of images need to be fused. Therefore, it is usually necessary to sort and register the obtained images first, and then perform image fusion.

[0065] During image fusion, the registered images to be fused are first imported. Multiple images to be fused are arranged sequentially according to the registration order. Typically, the first and second images to be fused are merged to obtain the first fused image. Then, the first fused image is merged with the third image to obtain the second fused image. Finally, the second fused image is merged with the fourth image to obtain the third fused image, and so on, until all images to be fused are fused, resulting in the final fused image. Since the fusion of two images to be fused and the fusion of one image to be fused with another image to be fused follow the same steps, both images to be fused will be referred to as "images to be fused" in this explanation.

[0066] Determine the boundary lines of the multiple images to be fused;

[0067] The shortest distance between each overlapping pixel in the overlapping region of the multiple images to be fused and the corresponding boundary line of the image to be fused is obtained respectively.

[0068] Determine the contribution weight of each image to be fused to each overlapping pixel;

[0069] Based on the contribution weight of each image to be fused to each overlapping pixel and the corresponding pixel value of each overlapping pixel in the corresponding image to be fused, the fusion pixel of each overlapping pixel is determined, and the fused image is obtained.

[0070] To further explain, determining the boundary lines of the multiple images to be fused specifically includes the following steps:

[0071] Perform grayscale processing on the images to be merged to obtain a grayscale image to be merged;

[0072] Add 1 to the gray value of each pixel in the grayscale image to be fused, and extend the boundary of the grayscale image to be fused by 1 pixel with a gray value of 0 to obtain the extended image to be fused.

[0073] Line detection is performed on the expanded image to be fused to obtain the boundary lines of the expanded image to be fused;

[0074] Subtract 1 from the coordinates of the endpoints of the boundary line of the expanded image to obtain the boundary line of the image to be fused.

[0075] To further explain, obtaining the shortest distance between each overlapping pixel in the overlapping region of multiple images to be fused and the corresponding boundary line of the image to be fused specifically includes the following steps:

[0076] Select any overlapping pixel in the overlapping region of the images to be fused;

[0077] Calculate the shortest distance between the overlapping pixel and the boundary line of the image to be fused;

[0078] Repeat the above steps until the shortest distance between each overlapping pixel in the overlapping region of the image to be fused and the boundary line of the image to be fused is obtained.

[0079] To further explain, determining the contribution weight of each image to be fused to each overlapping pixel based on the shortest distance specifically includes the following steps:

[0080] The obtained shortest distance is normalized to obtain the initial contribution weight of the image to be fused to each overlapping pixel.

[0081] A cosine transform is performed on each initial contribution weight to obtain the contribution weight of the image to be fused to each overlapping pixel.

[0082] To further explain, the process of obtaining the pixel value of each overlapping pixel in the corresponding image to be fused, and determining the fusion pixel for each overlapping pixel based on the contribution weight of each image to each overlapping pixel and the pixel value of each overlapping pixel in the corresponding image to be fused, to obtain the fused image, specifically includes the following steps:

[0083] Obtain the pixel value of any overlapping pixel in the overlapping region of the image to be fused;

[0084] The fused pixel value of the overlapping pixel is obtained by summing the products of the pixel value of the overlapping pixel in the image to be fused and the contribution weight of the image to be fused to the overlapping pixel.

[0085] Repeat the above steps until the fused pixel values ​​of all overlapping pixels in the overlapping area of ​​the images to be fused are obtained, thus completing the fusion of the images to be fused.

[0086] Example 2

[0087] Unlike Example 1, determining the boundary lines of the multiple images to be fused specifically includes the following steps:

[0088] The images to be fused are reduced by a reduction factor n, where n > 1;

[0089] For details, please refer to Figure 2 and Figure 3 Select two images to be merged, such as image a and image b.

[0090] The images to be merged, a and b, are reduced by a reduction factor n, for example, n=0.2. The size of both images to be merged is 3840*2160 pixels, thus obtaining the reduced image to be merged A and the reduced image to be merged B, both of which are 768*432 pixels in size.

[0091] By reducing the size of the images to be fused, the image size is reduced, the difficulty of subsequent image information processing is reduced, and the processing efficiency is improved.

[0092] The reduced image to be fused is processed into grayscale to obtain a grayscale image to be fused.

[0093] Specifically, the scaled-down images A and B to be merged are processed into grayscale, and then converted into two grayscale images to be merged using existing methods.

[0094] Add 1 to the gray value of each pixel in the grayscale image to be fused, and extend the boundary of the grayscale image to be fused by 1 pixel with a gray value of 0 to obtain the extended image to be fused.

[0095] Specifically, firstly, the grayscale value of each pixel in the two grayscale images to be merged is incremented by 1, so that the grayscale values ​​of the merged images are not all zero. Then, the outermost layer of the two grayscale images to be merged, with the grayscale value incremented by 1, is extended outward by 1 pixel with a grayscale value of 0. At this point, the outer edges of the top, bottom, left, and right boundaries of the two grayscale images to be merged are all wrapped with a pixel with a grayscale value of 0, and its length / width corresponds to the size of the corresponding grayscale image to be merged plus 1 pixel, thus obtaining the extended image to be merged.

[0096] Line detection is performed on the expanded image to be fused to obtain the boundary lines of the expanded image to be fused;

[0097] Specifically, the two expanded images to be fused are then binarized, that is, pixels with a gray level of 0 are processed to 0, and pixels with a gray level of any gray value from 1 to 255 are processed to 1. Then, the boundary lines of the two expanded images to be fused can be obtained quickly and accurately by Hough line detection.

[0098] Subtract 1 from the coordinates of the endpoints of the expanded boundary line of the image to be fused to obtain the reduced boundary line of the image to be fused.

[0099] Specifically, straight line detection is performed on the two expanded images to be merged, and the boundary lines of the two expanded images to be merged are obtained. The boundary lines at this time are the lines formed by the positions of the pixels with grayscale of 0 added during the expansion.

[0100] After obtaining the boundary lines of the two expanded images to be fused, it is necessary to subtract 1 from the coordinates of both ends of each boundary line of the expanded images to be fused. In two-dimensional coordinates, it is necessary to subtract 1 from both the x and y coordinates of the endpoints. If the coordinate value is negative, the coordinate value is added by 1. For ease of calculation, the expanded images to be fused are usually located in the first quadrant of the two-dimensional coordinates.

[0101] This yields the boundary lines of the reduced image A and the reduced image B.

[0102] To further explain, obtaining the shortest distance between each overlapping pixel in the overlapping region of multiple images to be fused and the corresponding boundary line of the image to be fused specifically includes the following steps:

[0103] Select any overlapping pixel in the overlapping region of the images to be fused;

[0104] Calculate the shortest distance between the overlapping pixel and the boundary line of the image to be fused.

[0105] Repeat the above steps until the shortest distance between each overlapping pixel in the overlapping region of the image to be fused and the boundary line of the image to be fused is obtained.

[0106] Specifically, select an overlapping pixel in the overlapping area of ​​the reduced images A and B. This overlapping pixel is pixel P in the reduced image A and pixel p in the reduced image B, and pixel P and pixel p have the same coordinates.

[0107] Then, calculate the minimum distance from the overlapping pixel to the boundary line of the scaled-down image A and the scaled-down image B. Usually, the image to be merged is a rectangle, and it will also be a rectangle after scaling down. At this time, calculate the distances from the overlapping pixel to the four boundary lines of the scaled-down image A as DA1, DA2, DA3 and DA4 respectively. Then, select the shortest distance from the four distances as the shortest distance DA from the overlapping pixel to the boundary line of the scaled-down image A.

[0108] If the foot of the perpendicular from the overlapping pixel to the four boundary lines of the reduced image A to be merged lies on the corresponding boundary line, then the shortest distance from the overlapping pixel to the corresponding boundary line is the perpendicular distance. If the foot of the perpendicular from the overlapping pixel to the four boundary lines of the reduced image A to be merged lies outside the corresponding boundary line, then the distance from the overlapping pixel to the endpoint of the corresponding boundary line is the shortest distance.

[0109] Similarly, the shortest distance DB from the overlapping pixel to the boundary line of the image B to be fused is calculated.

[0110] This process is repeated until the shortest distances DA1~DAk from each overlapping pixel in the overlapping region of the reduced-size images A and B to the reduced-size images A, and the shortest distances DB1~DBk from each overlapping pixel to the reduced-size images B are obtained, where k is the number of overlapping pixels in the overlapping region of the reduced-size images A and B, and k>1.

[0111] To further explain, determining the contribution weight of each image to be fused to each overlapping pixel specifically includes the following steps:

[0112] The obtained shortest distance is normalized to obtain the initial contribution weight of the image to be fused to each overlapping pixel.

[0113] Specifically, DAi and DBi are normalized to obtain the initial contribution weight value Norm_WAi of the reduced image A to the i-th overlapping pixel and the initial contribution weight value Norm_WBi of the reduced image B to the i-th overlapping pixel.

[0114] Norm_WAi=(DAi+least) / (DAi+DBi+2least)

[0115] Norm_WBi=1- Norm_WAi

[0116] Where i is the i-th overlapping pixel among k overlapping pixels, least is a constant. To avoid the denominator being 0, least can be selected as 0.000001 or other values ​​that are close to 0 but not equal to 0.

[0117] Thus, the initial contribution weights Norm_WA1~Norm_WAk of ​​the reduced image A to be fused to each overlapping pixel and the initial contribution weights Norm_W1~Norm_WBk of the reduced image B to be fused to each overlapping pixel are obtained.

[0118] Perform a cosine transform on each initial contribution weight to obtain the cosine contribution weight of the reduced image to be fused to each overlapping pixel.

[0119] Specifically, in order to reduce motion artifacts caused by the fusion of moving objects in the overlapping area, the initial contribution weights are processed by cosine transformation to reduce the impact of small weights on the fusion and reduce the generation of motion artifacts, thus obtaining cosine contribution weights.

[0120] For example, performing a cosine transform on the initial contribution weight value Norm_WAi yields the cosine contribution weight WAI, and performing a cosine transform on the initial contribution weight value Norm_WBi yields the cosine contribution weight WBi:

[0121] WAi=1.0-0.5 (cos(π×Norm_WAi)+1)

[0122] WBi=1-WAi

[0123] Thus, cosine transformation is performed on each initial contribution weight value to obtain cosine contribution weights WA1~WAk and cosine contribution weights WB1~WBk.

[0124] The cosine contribution weight of the image to be fused to each actual overlapping pixel is obtained by magnifying and reducing the image to be fused to each overlapping pixel by a magnification factor of 1 / n.

[0125] Specifically, since both the scaled-down images A and B are scaled-down images, the resulting cosine contribution weights are also the cosine contribution weights of the scaled-down images to the pixels in the overlapping region of the scaled-down images. Therefore, each cosine contribution weight is amplified by a magnification factor of 1 / n using bilinear interpolation to obtain the contribution weights WA1~WAk / n of image a to the pixels in the overlapping region of image a and image b, and the contribution weights WB1~WBk / n of image b to the pixels in the overlapping region of image a and image b.

[0126] The contribution weight of the image to be fused to each actual overlapping pixel.

[0127] To further explain, the process of determining the fusion pixel for each overlapping pixel, based on the contribution weight of each image to be fused to each overlapping pixel and the pixel value of each overlapping pixel in the corresponding image to be fused, and obtaining the fused image, specifically includes the following steps:

[0128] Obtain the pixel value of any overlapping pixel in the overlapping region of the image to be fused;

[0129] The fused pixel value of the overlapping pixel is obtained by summing the products of the pixel value of the overlapping pixel in the image to be fused and the contribution weight of the image to be fused to the overlapping pixel.

[0130] Repeat the above steps until the fused pixel values ​​of all overlapping pixels in the overlapping area of ​​the images to be fused are obtained, thus completing the fusion of the images to be fused.

[0131] Specifically, obtain any overlapping pixel in the overlapping region of the images to be merged, such as the I-th overlapping pixel. At this time, the pixel value of the I-th overlapping pixel in the image to be merged is PI, and the pixel value of the I-th pixel in the image to be merged is pI.

[0132] Calculate the pixel value O of the fused i-th pixel:

[0133] O = WAI × PI + WBI × pI

[0134] I represents the i-th overlapping pixel among the k / n overlapping pixels in the overlapping region of images a and b to be merged; WAI represents the contribution weight of image a to the i-th overlapping pixel in the overlapping region of images a and b to be merged; and WBI represents the contribution weight of image b to the i-th overlapping pixel in the overlapping region of images a and b to be merged.

[0135] refer to Figure 4 This process calculates the fused pixel value of the I-th overlapping pixel in the overlapping area of ​​the two images a and b. The calculation is repeated for each overlapping pixel in the overlapping area of ​​the two images a and b to obtain the fused pixel value. This allows each overlapping pixel in the overlapping area of ​​the two images a and b to be displayed according to the fused pixel value, thus completing the fusion of the two images.

[0136] In this way, the fusion of the merged image and the third image to be merged begins, until the fusion of all images is completed. The operation is simple and convenient, the fusion efficiency is high, and it meets the needs of practical use.

[0137] In contrast, such as Figure 5 The image obtained by fusing image a and image b using existing methods has poor fusion results and cannot meet practical needs.

[0138] Example 3

[0139] refer to Figure 6 and Figure 7 This embodiment also provides a fast image fusion system based on field of view information, including:

[0140] The image acquisition unit is used to acquire multiple registered images to be merged.

[0141] The image boundary determination unit is used to determine the boundary lines of multiple images to be merged respectively;

[0142] The shortest distance calculation unit between overlapping pixels and edge lines is used to obtain the shortest distance between each overlapping pixel in the overlapping region of multiple images to be fused and the corresponding boundary line of the image to be fused.

[0143] The overlapping pixel contribution weight calculation unit is used to determine the contribution weight of each image to be fused to each overlapping pixel.

[0144] The image fusion unit is used to obtain the pixel value of each overlapping pixel in the corresponding image to be fused, and to determine the fusion pixel of each overlapping pixel based on the contribution weight of each image to be fused to each overlapping pixel and the pixel value of each overlapping pixel in the corresponding image to be fused, thereby obtaining the fused image. The specific steps include:

[0145] Further specifying, the edge determination unit of the image to be fused includes:

[0146] The image reduction module is used to reduce the image to be merged according to a coefficient n, where n > 1;

[0147] The grayscale image to be merged module is used to perform grayscale processing on the reduced image to be merged to obtain the grayscale image to be merged.

[0148] The extended image acquisition module is used to add 1 to the gray value of each pixel in the grayscale image to be merged, and to extend the boundary of the grayscale image to be merged by 1 pixel with a gray value of 0 to obtain the extended image to be merged.

[0149] The module for obtaining the boundary line of the expanded image to be fused is used to perform line detection on the expanded image to be fused to obtain the boundary line of the expanded image to be fused.

[0150] The module for obtaining the boundary line of the image to be fused by reducing the coordinate value of the endpoint of the boundary line of the expanded image to be fused by 1 is used to obtain the boundary line of the image to be fused by reducing the boundary line.

[0151] Further specifying, the unit for calculating the shortest distance between overlapping pixels and the edge line includes:

[0152] The module for calculating the distance between overlapping pixels and the image to be fused is used to calculate the shortest distance between each overlapping pixel and the boundary line of the image to be fused.

[0153] Further specifying, the overlapping pixel contribution weight calculation unit includes:

[0154] The initial contribution weight calculation module is used to normalize the shortest distance corresponding to each overlapping pixel to obtain the initial contribution weight of the reduced image to be fused to each overlapping pixel.

[0155] The cosine contribution weight calculation module is used to perform a cosine transformation on each initial contribution weight to obtain the cosine contribution weight of the reduced image to be fused to each overlapping pixel.

[0156] The contribution weight calculation module is used to magnify and reduce the cosine contribution weight of the image to be fused to each overlapping pixel by a magnification factor of 1 / n, so as to obtain the contribution weight of the image to be fused to each overlapping pixel.

[0157] Further specifying, the image fusion unit includes:

[0158] The overlapping pixel acquisition module is used to acquire the pixel values ​​of overlapping pixels in the overlapping region of the image to be fused.

[0159] The overlapping pixel fusion module is used to obtain the fusion pixel value of each overlapping pixel in the overlapping region of the image to be fused by summing the product of the pixel value of each overlapping pixel in the overlapping region of the image to be fused and the contribution weight of the image to be fused to that overlapping pixel in the image to be fused, thereby completing the fusion of the images to be fused.

[0160] While this application provides method operation steps as shown in the embodiments or flowcharts, more or fewer operation steps may be included based on conventional or non-inventive labor. The order of steps listed in this embodiment is merely one possible execution order among many and does not represent the only execution order. In actual device or client product execution, the method can be executed sequentially or in parallel according to the method shown in this embodiment or the accompanying drawings.

[0161] The above embodiments are only used to illustrate the technical solutions of this application, and are not intended to limit this application. Although this application has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the foregoing embodiments, or make equivalent substitutions for some or all of the technical features therein; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of this application.

Claims

1. A fast image fusion method based on field-of-view information, characterized in that, Includes the following steps: Acquire multiple registered images to be fused; Determine the boundary lines of the multiple images to be fused; The method involves obtaining the shortest distance between each overlapping pixel in the overlapping region of multiple images to be merged and the corresponding boundary line of the image to be merged; specifically, obtaining the shortest distance between each overlapping pixel in the overlapping region of multiple images to be merged and the corresponding boundary line of the image to be merged includes the following steps: Select any overlapping pixel in the overlapping region of the images to be fused; calculate the shortest distance between the overlapping pixel and the boundary line of the images to be fused; repeat the above steps until the shortest distances between all overlapping pixels in the overlapping region of the images to be fused and the boundary line of the images to be fused are obtained; The contribution weight of each image to be fused to each overlapping pixel is determined based on the shortest distance; wherein, determining the contribution weight of each image to be fused to each overlapping pixel specifically includes the following steps: The shortest distance is normalized to obtain the initial contribution weight of the image to be fused to each overlapping pixel; a cosine transform is performed on each initial contribution weight to obtain the contribution weight of the image to be fused to each overlapping pixel. Obtain the pixel value of each overlapping pixel in the corresponding image to be fused. Based on the contribution weight of each image to be fused to each overlapping pixel and the pixel value of each overlapping pixel in the corresponding image to be fused, determine the fusion pixel of each overlapping pixel to obtain the fused image.

2. The image fast fusion method based on field of view information according to claim 1, characterized in that, The specific steps for determining the boundary lines of the multiple images to be fused include: Perform grayscale processing on the images to be merged to obtain a grayscale image to be merged; Add 1 to the gray value of each pixel in the grayscale image to be fused, and extend the boundary of the grayscale image to be fused by 1 pixel with a gray value of 0 to obtain the extended image to be fused. Line detection is performed on the expanded image to be fused to obtain the boundary lines of the expanded image to be fused; Subtract 1 from the coordinates of the endpoints of the boundary line of the expanded image to obtain the boundary line of the image to be fused.

3. The image fast fusion method based on field of view information according to claim 1, characterized in that, The process of determining the fusion pixel for each overlapping pixel based on the contribution weight of each image to be fused to each overlapping pixel and the corresponding pixel value of each overlapping pixel in the corresponding image to be fused, and obtaining the fused image, specifically includes the following steps: Obtain the pixel value of any overlapping pixel in the overlapping region of the image to be fused; The fused pixel value of the overlapping pixel is obtained by summing the products of the pixel value of the overlapping pixel in the image to be fused and the contribution weight of the image to be fused to the overlapping pixel. Repeat the above steps until the fused pixel values ​​of all overlapping pixels in the overlapping area of ​​the images to be fused are obtained, thus completing the fusion of the images to be fused.

4. A fast image fusion system based on field-of-view information, characterized in that, include: The image acquisition unit is used to acquire multiple registered images to be merged. The image boundary determination unit is used to determine the boundary lines of multiple images to be merged respectively; The shortest distance calculation unit between overlapping pixels and edge lines is used to obtain the shortest distance between each overlapping pixel in the overlapping region of multiple images to be fused and the corresponding boundary line of the image to be fused. The shortest distance calculation unit between overlapping pixels and the edge line includes: a distance calculation module between overlapping pixels and the image to be fused, used to calculate the shortest distance between each overlapping pixel and the boundary line of the image to be fused; The overlapping pixel contribution weight calculation unit is used to determine the contribution weight of each image to be fused to each overlapping pixel based on the shortest distance. The overlapping pixel contribution weight calculation unit includes: an initial contribution weight calculation module, used to normalize the shortest distance corresponding to each overlapping pixel to obtain the initial contribution weight of the image to be fused to each overlapping pixel; and a contribution weight calculation module, used to perform a cosine transform on each initial contribution weight to obtain the cosine contribution weight of the image to be fused to each overlapping pixel. The image fusion unit is used to obtain the pixel value of each overlapping pixel in the corresponding image to be fused, and determine the fusion pixel of each overlapping pixel based on the contribution weight of each image to be fused to each overlapping pixel and the corresponding pixel value of each overlapping pixel in the corresponding image to be fused, so as to obtain the fused image.

5. The image fast fusion system based on field of view information according to claim 4, characterized in that, The image edge determination unit includes: The grayscale image to be merged module is used to perform grayscale processing on the image to be merged to obtain the grayscale image to be merged. The extended image acquisition module is used to add 1 to the gray value of each pixel in the grayscale image to be merged, and to extend the boundary of the grayscale image to be merged by 1 pixel with a gray value of 0 to obtain the extended image to be merged. The module for obtaining the boundary line of the expanded image to be fused is used to perform line detection on the expanded image to be fused to obtain the boundary line of the expanded image to be fused. The module for obtaining the boundary line of the image to be fused is used to subtract 1 from the coordinate values ​​of the endpoints of the extended boundary line of the image to be fused to obtain the boundary line of the image to be fused.

6. The image fast fusion system based on field of view information according to claim 4, characterized in that, The image fusion unit includes: The overlapping pixel acquisition module is used to acquire the pixel values ​​of overlapping pixels in the overlapping region of the image to be fused. The overlapping pixel fusion module is used to obtain the fusion pixel value of each overlapping pixel in the overlapping region of the image to be fused by summing the product of the pixel value of each overlapping pixel in the overlapping region of the image to be fused and the contribution weight of the image to be fused to that overlapping pixel in the image to be fused, thereby completing the fusion of the images to be fused.

Images