Measurement method and device based on multi-view slit full-length X-ray photography

Abstract

The invention discloses a measurement method based on multi-view slit full-length X-ray photography, and the method comprises the steps: shooting a to-be-shot object through a plurality of shooting assemblies, moving the to-be-shot object according to a preset moving speed, and shooting the to-be-shot object according to a preset frame rate, thereby obtaining a plurality of image pictures; and selecting a group of image pictures, and carrying out image splicing. And if the height of the flat plate is smaller than the thickness of the parallel light, splicing the adjacent image pictures according to the physical position of the corresponding flat plate to obtain a spliced image. Otherwise, image matching is carried out, the optimal transverse offset and the optimal longitudinal offset between the adjacent image pictures are obtained, and a spliced image is obtained. And mapping the vertical coordinates and the horizontal coordinates of all the spliced images to real physical position pixels to realize measurement based on the matched spliced images. According to the invention, continuous work of a high-frame-rate and high-power radiation bulb tube is not needed. The method is suitable for narrow slits in any mode. And the method has certain tolerance to the motion error of the X-ray equipment. And extra workpiece shooting for calibration in the shooting process is not needed.

Description

Technical Field

[0001] The invention belongs to the field of medical imaging, and in particular relates to a measurement method and device based on multi-viewing angle slit full-length X-ray photography. Background Art

[0002] Traditional full-length human body X-ray photography uses the movement and swing of the X-ray source head and the movement of the flat-panel detector to capture multiple X-ray images. The range of a single X-ray image covers the entire range of the flat-panel detection, and the images are stitched together and provided to the physician for analysis. However, actual measurements have found that the measured length is significantly distorted, with errors reaching the centimeter level, and cannot reflect the actual patient's bone morphology information. In addition, it is necessary to place a fixed-size workpiece during shooting and perform depth correction to obtain accurate length information, such as placing coins on the body surface. Once the patient is photographed, it is impossible to perform post-correction again. In addition, the workpiece is located on the body surface and cannot reflect the actual physical depth of the bones. There are technical disadvantages such as inaccurate images and the need for correction during shooting.

[0003] In the existing technology, EOS imaging achieves the effect of linear narrow slit shooting and provides an accurate full-length measurement solution. It relies on mechanical precision to directly shoot objects linearly and perform high stacking, skipping the image stitching and fusion process of traditional images, and determining the lateral magnification by limiting the standing position. However, EOS imaging has extremely strict requirements on the physical accuracy of the frame. Once the up and down movement of the track cannot achieve the stitching accuracy, repeated stripes will appear in the image, making it difficult to achieve large-scale popular shooting. In addition, its pixel-by-pixel shooting requires high-power tubes to work continuously, and the overall design is complex, difficult, and costly. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a measurement method and device based on multi-view slit full-length X-ray photography to solve the problem that full-length X-ray photography of the human body is inaccurate and requires shooting correction.

[0005] To solve the above problems, the technical solution of the present invention is:

[0006] A measurement method based on multi-view slit full-length X-ray photography comprises the following steps:

[0007] Using several groups of shooting components to shoot the object to be shot, each group of shooting components includes a matching radiation source and a plate, each group of shooting components moves at their own preset moving speed, and shoots at their own preset frame rate, and each group of shooting components shoots a number of images;

[0008] Select a number of images taken by any group of shooting components, and stitch the images together, specifically:

[0009] Compare the plate height of the current shooting component with the thickness of the parallel light emitted by the radiation source and constrained by the beam splitter;

[0010] If the height of the plate is less than the parallel light thickness, the adjacent image pictures are spliced ​​according to the physical positions of the corresponding plates to obtain a spliced ​​image;

[0011] If the height of the plate is greater than the parallel light thickness, adjacent images are matched to obtain the best lateral offset and the best longitudinal offset between adjacent images, and all images are spliced ​​to obtain a spliced ​​image according to the best lateral offset and the best longitudinal offset;

[0012] After all the image pictures corresponding to all the shooting components are stitched together, all the vertical and horizontal coordinates of all the stitched images are mapped to the real physical position pixels.

[0013] Among them, if the height of the plate is less than the parallel light thickness, the following conditions are met:

[0014]

[0015] Among them, V is the preset moving speed of the shooting component, F is the preset shooting frame rate of the shooting component, and h is the parallel light thickness.

[0016] Specifically, for the flat panel of the shooting component, the use range ROI of the flat panel = [X, Y, width, height], where X and Y are the horizontal and vertical coordinates of the starting position of the use range of the flat panel, width is the width of the flat panel, and height is the height of the flat panel. The specific steps of matching and stitching adjacent image pictures are as follows:

[0017] Select two adjacent image pictures in order, namely the first image picture and the second image picture, search for the best horizontal offset and the best vertical offset of the second image picture relative to the first image picture, and traverse all possible horizontal offsets and vertical offsets, wherein the search range of the horizontal offset is 0 to width, and the search range of the vertical offset is 0 to height;

[0018] Assume that the lateral offset and longitudinal offset are selected as [x 1 ,y 1 ], take the valid image range from the first image as:

[0019] img1ROI=[x 1 ,height―y 1 ,width―x 1 ,y 1 ]

[0020] Take the valid image range from the second image as img2ROI=[0,0,x1 ,y 1 ]

[0021] Calculate the similarity between the first image and the second image, and the horizontal offset and the vertical offset corresponding to the highest similarity value are the optimal horizontal offset and the optimal vertical offset;

[0022] splicing the first image picture with the second image picture based on the optimal lateral offset and the optimal longitudinal offset;

[0023] Select subsequent adjacent image pictures and repeat the above steps until all the image pictures are spliced ​​together.

[0024] Among them, all the vertical coordinates of all the stitched images are mapped to the real physical position pixels:

[0025] The height of the stitched image satisfies the following formula:

[0026] HEIGHT=n*height―yoffset 1 ―...―yoffset n―1

[0027] n is the number of images in the stitched image sequence, traverse HEIGHT, such as any point y raw , the physical height corresponding to the real mapping of the stitched image is:

[0028] h map

[0029] =(idx-1)*V*(1 / F)+(V*(1 / F)) / (height-yoffset idx―1 )*(y raw

[0030] ―h idx―1 )

[0031] Among them, idx is the current query y raw is the sequence in which it is located, and the judgment method is to compare it with the accumulated height H stich Take the first non-positive value of the difference; (idx-1)*V*(1 / F) is y raw The actual physical height of the image in the previous (idx-1) sequence; (V*(1 / F)) / (height-yoffset idx―1 ) is y raw The average pixel physical height of the current sequence image; y raw ―h idx―1 for y raw The number of rows in the current sequence of images.

[0032] Among them, all the horizontal coordinates of all the stitched images are mapped to the real physical position pixels:

[0033] The depth of the object to be photographed is determined by multi-view imaging, and then the stitched image is mapped to all horizontal coordinates to satisfy the following formula:

[0034] X map =(x raw ―0.5WIDTH)*k*u

[0035] Where k is the magnification, which is calculated as k = D / D-d, D-d is the depth of the object to be photographed, D is the depth of the radiation source from the flat plate, and d is the distance from the object to be photographed to the flat plate; u is the physical size of each pixel of the flat plate; x raw Select the horizontal coordinate of a measurement point for the stitched image; WIDTH is the width of the stitched image.

[0036] Further preferably, the method further comprises calculating the depth of the object to be photographed: selecting a point P on the object to be photographed, the point P being imaged on at least two flat plates, the physical positions of the two flat plates being obtainable by a position change matrix T, the position change matrix T being composed of a rotation R and a displacement t: T=[R t];

[0037] Then the depth d of the object to be photographed relative to the two plates is 1 ,d 2 Satisfies the following formula:

[0038]

[0039] Among them, p 1 ,p 2 is a homogeneous coordinate, with coordinates [x, y, 1], where x and y are coordinates on the plate, and the depth d can be obtained by solving 1 ,d 2 .

[0040] A measuring device based on multi-view slit full-length X-ray photography, which realizes the above-mentioned measuring method based on multi-view slit full-length X-ray photography, comprises a plurality of longitudinal moving axes, a main control module, and a plurality of shooting components arranged on the longitudinal moving axes;

[0041] Each group of shooting components includes a radiation source and a flat plate. The radiation source and the flat plate in the same group are arranged on different longitudinal moving axes and are driven by the longitudinal moving axes to realize synchronous movement, and shoot the object to be shot to obtain a group of image pictures.

[0042] The main control module is configured to receive a number of images taken by all the shooting components, and stitch the images together, compare the plate height of the current shooting component with the thickness of the parallel light emitted by the radiation source and constrained by the beam splitter; if the plate height is less than the parallel light thickness, stitch the adjacent images together according to the physical positions of the corresponding plates to obtain a stitched image; if the plate height is greater than the parallel light thickness, match the adjacent images together to obtain the best lateral offset and the best longitudinal offset between the adjacent images, and stitch all the images together according to the best lateral offset and the best longitudinal offset to obtain a stitched image; after all the images corresponding to all the shooting components are stitched together, map all the ordinates and abscissas of all the stitched images to real physical position pixels.

[0043] A computer device includes a memory and a processor. The memory stores computer-readable instructions. When the computer-readable instructions are executed by the processor, the processor executes the above-mentioned measurement method based on multi-view slit full-length X-ray photography.

[0044] A storage medium storing computer-readable instructions, when the computer-readable instructions are executed by one or more processors, enables the one or more processors to execute the above-mentioned measurement method based on multi-view slit full-length X-ray photography.

[0045] Due to the adoption of the above technical solution, the present invention has the following advantages and positive effects compared with the prior art:

[0046] The present invention supports low frame rate sequential pulse shooting, and does not require high frame rate and high power radiation tube to work continuously.

[0047] The present invention can support any pattern of narrow slits, and the narrower the narrow slit, the higher the precision.

[0048] The present invention is compatible with the conventional full-length image shooting, splicing and fusion process and has a certain tolerance for the motion error of the X-ray equipment.

[0049] The present invention supports multi-view shooting and can self-correct depth information, without the need to additionally shoot workpieces for calibration during the shooting process. BRIEF DESCRIPTION OF THE DRAWINGS

[0050] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the following detailed description of the preferred embodiment.The drawings are only for the purpose of illustrating the preferred embodiments and are not to be construed as limiting the invention.

[0051] Figure 1 A flow chart of a measurement method based on multi-view slit full-length X-ray photography of the present invention;

[0052] Figure 2This is a structural diagram of a measuring device based on multi-view slit full-length X-ray photography of the present invention;

[0053] Figure 3 It is a shooting schematic diagram of the shooting assembly of the present invention;

[0054] Figure 4 A schematic diagram of image stitching under physical stacking of the present invention;

[0055] Figure 5 A schematic diagram of image stitching under image matching of the present invention;

[0056] Figure 6 It is a schematic diagram of obtaining the depth of an object to be photographed according to the present invention. DETAILED DESCRIPTION

[0057] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the specific implementation methods of the present invention will be described below with reference to the accompanying drawings. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings and other implementation methods can be obtained based on these drawings without creative work.

[0058] In order to simplify the drawings, only the parts related to the present invention are schematically shown in each figure, and they do not represent the actual structure of the product. In addition, in order to simplify the drawings and facilitate understanding, in some figures, only one of the parts with the same structure or function is schematically drawn or marked. In this article, "one" not only means "only one", but also means "more than one".

[0059] The following is a further detailed description of a measurement method and device based on multi-view slit full-length X-ray photography proposed by the present invention in conjunction with the accompanying drawings and specific embodiments. The advantages and features of the present invention will become more apparent from the following description and claims.

[0060] Example

[0061] See also Figures 1 to 6 This embodiment provides a measurement method and device based on multi-view slit full-length X-ray photography. Since the implementation of this embodiment relies on the device, refer to Figure 2, the measuring device based on multi-view slit full-length X-ray photography is now described: it includes a number of longitudinal moving axes, a main control module, and a number of shooting components arranged on the longitudinal moving axes (i.e., a frame). Among them, there are at least 4 longitudinal moving axes, that is, 2 groups of shooting components can be set, and the objects to be photographed are located in the shooting areas of the 2 groups of shooting components. Each group of shooting components includes a radiation source and a flat-panel detector. The radiation source and the flat-panel detector of the same group are arranged on different longitudinal moving axes and are arranged opposite to each other. They are driven by the two longitudinal moving axes to achieve synchronous movement. During the movement, the objects to be photographed are photographed to obtain a group of image pictures. The flat-panel detector is a dynamic detector, and its partial area or the entire area can be selected for use.

[0062] The main control module can receive several images taken by all shooting components, and stitch the images of the same group. By comparing the plate ROI height of the current shooting component with the thickness of the parallel light emitted by the radiation source through the beam-beamer constraint, the stitching can be divided into two cases. First, if the plate ROI height is less than the parallel light thickness, the adjacent image pictures are stitched according to the physical position of the corresponding plate to obtain a stitched image. Second, if the plate ROI height is greater than the parallel light thickness, the adjacent image pictures are matched to obtain the best lateral offset and the best longitudinal offset between the adjacent image pictures, and all the image pictures are stitched together according to the best lateral offset and the best longitudinal offset to obtain a stitched image. Finally, after all the image pictures corresponding to all the shooting components are stitched together, all the ordinates and abscissas of all the stitched images are mapped to the real physical position pixels to achieve measurement based on matching stitched images, that is, the image pictures correspond to the actual coordinates, and the patient's lesions or opening positions can be clearly obtained.

[0063] See also Figure 1 , using the above-mentioned measurement device based on multi-view slit full-length X-ray photography, the corresponding method is now described, which specifically includes the following steps:

[0064] First, multiple groups of shooting components are used to shoot the object to be shot. Each group of shooting components moves at its own preset moving speed and shoots from top to bottom at its own preset frame rate. Specifically, assuming that the moving speed of the longitudinal moving axis is V, different pairs of radiation sources and flat panels can move at different speeds V = {V 1 , V 2 , V 3 ...}movement; Assuming the frame rate of the flat panel is F, different pairs of radiation sources and flat panels can be photographed at different frame rates F = {F 1 , F 2 , F 3……}. The tablet's usage range is ROI = [X, Y, width, height], where X and Y are the horizontal and vertical coordinates of the starting position of the tablet's usage range, width is the width of the tablet, and height is the height of the tablet. The ROI of different tablets can be different. ROI = {ROI 1 , ROI 2 , ROI 3 ……}. After shooting from top to bottom at different heights, each shooting component takes a series of images: radiation source 1 and plate 1: Images1 = {img 1 ,img 2 ,…img n}; Radiation source 2 and plate 2: Images2 = {img 1 ,img 2 ,…img m}; ...radiation source N and plate N: ImagesN = {img 1 ,img 2 ,…img k Due to the different distances traveled and the shooting frame rates of different pairs of plates and radiation sources, the number of images obtained is also different.

[0065] Next, the narrow image areas of different pairs of images are stitched together. Since the height of the flat ROI may be too small, the conventional method cannot be used to find the feature points of the adjacent images in the image sequence for matching and stitching. The pixel content of the image should be used as much as possible to match the physical positions of the front and back images, and then stitch the images.

[0066] Select a series of images taken by any group of shooting components, and stitch the above images together. The sequence of the image images is set as: Images1 = {img 1 ,img 2 ,…img n}. Compare the height of the flat plate ROI of the current shooting component with the thickness of the parallel light emitted by the radiation source and constrained by the beam splitter, such as Figure 3 and Figure 4 As shown in the figure, if the height of the flat plate ROI is less than the parallel light thickness, there is no need to perform a matching search for the upper and lower adjacent images. The adjacent image img 1 toimg n The corresponding panels are stitched together according to their physical positions to obtain a stitched image. The following conditions must be met at the same time:

[0067]

[0068] Among them, V is the preset moving speed of the shooting component, F is the preset shooting frame rate of the shooting component, and h is the parallel light thickness.

[0069] by Figure 5 For example, if the height of the flat plate ROI is greater than the parallel light thickness, the adjacent images are matched to obtain the best horizontal offset and the best vertical offset between the adjacent images. All images are stitched together according to the best horizontal offset and the best vertical offset to obtain a stitched image. The image matching process is as follows: For img 1 and to img 2 , the goal is to search for img 2 Relative to img 1 The optimal horizontal offset xoffset and the optimal vertical offset yoffset are searched. The search process is to traverse all possible horizontal offsets and vertical offsets, where the search range of the horizontal offset is the width of img, that is, 0~width, and the search range of the vertical offset is the height of img, that is, 0~height. Assume that the horizontal offset and vertical offset are selected as [x 1 ,y 1 ], from img 1 The corresponding valid image range is:

[0070] img 1 ROI = [x 1 ,height―y 1 ,width―x 1 ,y 1 ]

[0071] From img 2 Take the valid image range as img 2 ROI=[0,0,x 1 ,y 1 ]

[0072] That is, take the overlapping part of the two. Then, calculate img 1 with img 2 The horizontal offset and vertical offset corresponding to the highest similarity value are the optimal horizontal offset and the optimal vertical offset. The first image and the second image are spliced ​​together based on the optimal horizontal offset and the optimal vertical offset. 2 with img 3 ,img 3 with img 4 , …), repeat the above steps until all the images are stitched together.

[0073] There is no unique way to calculate similarity. For example, the commonly used NCC (normalized cross correlation) is calculated as follows:

[0074]

[0075] Among them, μ I for img 1 The mean value, μ T for img 2 The mean of (x, y) is the coordinate of the image. If we get img 1 ROI and img 2 The similarity of ROI is s 1 , then a series of possible similarity values ​​S = {s 1 ,…,s n}, assuming the best horizontal offset xoffset under the best similarity of the group 1 With the optimal vertical offset yoffset 1 Then, the two images are spliced ​​according to the offset to obtain another longer image, and the current accumulated height h is recorded. 2 =2*height―yoffset 1 .

[0076] In all images Images1 = {img 1 ,img 2 ,…img n}After stitching in sequence, a stitched image is obtained, and its corresponding stitching result is OFFSET = {[xyoffset 1 ,yoffset 1 ],…,[xyoffset n―1 ,yoffset n―1 ]}, the cumulative height is h stich ={h 1 ,……,h n}, where h 1 It is the sequence height.

[0077] After all the images corresponding to all the shooting components are stitched together, all the vertical and horizontal coordinates of all the stitched images are mapped to the real physical position pixels to achieve measurement based on matching stitched images.

[0078] Specifically, all the vertical coordinates of a series of stitched images are mapped to real physical position pixels, and the height corresponding to the stitched image is HEIGHT = n*height-yoffset 1 ―...―yoffset n―1 .

[0079] Where n is the number of images in the stitched image sequence, and height is the height of the image sequence. Traverse HEIGHT and select any point y raw , the physical height corresponding to its real mapping is:

[0080] h map

[0081] =(idx-1)*V*(1 / F)+(V*(1 / F)) / (height-yoffset idx―1 )*(y raw

[0082] ―h idx―1 )

[0083] Among them, idx is the current query y raw The sequence is determined by the cumulative height H stich Take the first non-positive value of the difference; (idx-1)*V*(1 / F) is y raw The actual physical height of the rack that the image of the previous (idx-1) sequence passes through; (V*(1 / F)) / (height-yoffset idx―1 ) is y raw The average pixel physical height of the current sequence image; y raw ―h idx―1 for y raw The number of rows in the current sequence of images. Then, all the horizontal coordinates of all the stitched images are mapped to the real physical position pixels. The mapping method can determine the depth D-d of the object to be photographed through multi-view images, and then map all the horizontal coordinates of the stitched images to meet the following formula:

[0084] X map =(x raw ―0.5WIDTH)*k*u

[0085] Where k is the magnification, which is calculated as k = D / D-d, D-d is the depth of the object to be photographed, D is the depth of the radiation source from the flat plate, and d is the distance from the object to be photographed to the flat plate; u is the physical size of each pixel of the flat plate; x raw Select the horizontal coordinate of a measurement point for the stitched image; WIDTH is the width of the stitched image.

[0086] See also Figure 6 Preferably, the present embodiment further includes calculating the depth of the object to be photographed, that is, calculating the depth in the horizontal coordinate mapping. A point P on the object to be photographed is selected. The point P can be a human anatomical point or a photographing workpiece, or a fixed point on the photographing system, and can be imaged at the same time as the object to be photographed. In addition, the point P is imaged on at least two flat panels to Figure 6For example, assuming that the point P can be imaged on imaging plates 1 and 2, and its projections are P' and P", assuming a pair of radiation source 1 and plate 1 and a pair of radiation source 2 and plate 2, their physical positions can be obtained by a position change matrix T, which consists of rotation R and displacement t: T = [R t].

[0087] Then the depth of the object to be photographed relative to the two plates is d 1 ,d 2 , and satisfies the following formula:

[0088]

[0089] Among them, p 1 ,p 2 is a homogeneous coordinate, that is, the coordinate is [x, y, 1], where x and y are the coordinates on the plate. Further solving can get the depth d 1 ,d 2 .

[0090] Furthermore, this embodiment also provides a computer device, including a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the measurement method based on multi-view slit full-length X-ray photography as described above.

[0091] More preferably, this embodiment also provides a storage medium storing computer-readable instructions, which, when executed by one or more processors, enables the one or more processors to execute the measurement method based on multi-view slit full-length X-ray photography as described above.

[0092] The embodiments of the present invention are described in detail above in conjunction with the accompanying drawings, but the present invention is not limited to the above embodiments. Even if various changes are made to the present invention, if these changes fall within the scope of the claims of the present invention and their equivalents, they still fall within the protection scope of the present invention.

Claims

1. A measurement method based on multi-view slit full-length X-ray photography, It is characterized in that The steps include: Using several groups of shooting components to shoot the object to be shot, each group of shooting components includes a matching radiation source and a plate, each group of shooting components moves at their own preset moving speed, and shoots at their own preset frame rate, and each group of shooting components shoots a number of images; Select a number of images taken by any group of shooting components, and stitch the images together, specifically: Compare the plate height of the current shooting component with the thickness of the parallel light emitted by the radiation source and constrained by the beam splitter; If the height of the plate is less than the parallel light thickness, the adjacent image pictures are spliced ​​according to the physical positions of the corresponding plates to obtain a spliced ​​image; If the height of the plate is greater than the parallel light thickness, adjacent images are matched to obtain the best lateral offset and the best longitudinal offset between adjacent images, and all images are spliced ​​to obtain a spliced ​​image according to the best lateral offset and the best longitudinal offset; After all the image pictures corresponding to all the shooting components are stitched together, all the vertical and horizontal coordinates of all the stitched images are mapped to the real physical position pixels.

2. The measurement method based on multi-view slit full-length X-ray photography according to claim 1, It is characterized in that If the plate height is less than the parallel light thickness, the following conditions are met: Among them, V is the preset moving speed of the shooting component, F is the preset shooting frame rate of the shooting component, and h is the parallel light thickness.

3. The measurement method based on multi-view slit full-length X-ray photography according to claim 1, It is characterized in that For the flat panel of the shooting component, the use range ROI of the flat panel = [X, Y, width, height], where X and Y are the horizontal and vertical coordinates of the starting position of the use range of the flat panel, width is the width of the flat panel, and height is the height of the flat panel. The specific steps for matching and stitching adjacent images are as follows: Select two adjacent image pictures in order, namely the first image picture and the second image picture, search for the best horizontal offset and the best vertical offset of the second image picture relative to the first image picture, and traverse all possible horizontal offsets and vertical offsets, wherein the search range of the horizontal offset is 0 to width, and the search range of the vertical offset is 0 to height; Assume that the lateral offset and longitudinal offset are selected as [x 1 ,y 1 ], the effective image range is taken from the first image: img1ROI=[x 1 ,height-y 1 ,width-x 1 ,y 1 ] Take the valid image range from the second image as img2ROI=[0, 0, x 1 ,y 1 ] Calculate the similarity between the first image and the second image, and the horizontal offset and the vertical offset corresponding to the highest similarity value are the optimal horizontal offset and the optimal vertical offset; splicing the first image picture with the second image picture based on the optimal lateral offset and the optimal longitudinal offset; Select subsequent adjacent image pictures and repeat the above steps until all the image pictures are spliced ​​together.

4. The measurement method based on multi-view slit full-length X-ray photography according to claim 3, It is characterized in that Map all vertical coordinates of all stitched images to real physical location pixels: The height of the stitched image satisfies the following formula: HEIGHT=n*height-yoffset 1 -...-yoffset n-1 n is the number of images in the stitched image sequence, traverse HEIGHT, such as any point y raw , the physical height corresponding to the real mapping of the stitched image is: h map =(idx-1)*V*(1 / F)+(V*(1 / F)) / (height-yoffset idx-1 )*(y raw -h idx-1 ) Among them, idx is the current query y raw is the sequence in which it is located, and the judgment method is to compare it with the accumulated height H stich Take the first non-positive value of the difference; (idx-1)*V*(1 / F) is y raw The actual physical height of the image in the previous (idx-1) sequence; (V*(1 / F)) / (height-yoffset idx-1 ) is y raw The average pixel physical height of the current sequence image; y raw -h idx-1 for y raw The number of rows in the current sequence of images.

5. The measurement method based on multi-view slit full-length X-ray photography according to claim 4, It is characterized in that Map all horizontal coordinates of all stitched images to real physical location pixels: The depth of the object to be photographed is determined by multi-view imaging, and then the stitched image is mapped to all horizontal coordinates to satisfy the following formula: X map =(x raw -0.5WIDTH)*k*u Where k is the magnification, which is calculated as k = D / Dd, Dd is the depth of the object to be photographed, D is the depth of the radiation source from the flat plate, d is the distance from the object to be photographed to the flat plate; u is the physical size of each pixel of the flat plate; x raw Select the horizontal coordinate of a measurement point for the stitched image; WIDTH is the width of the stitched image.

6. The measurement method based on multi-view slit full-length X-ray photography according to claim 5, It is characterized in that The method also includes calculating the depth of the object to be photographed: selecting a point P on the object to be photographed, the point P is imaged on at least two flat plates, and the physical positions of the two flat plates can be obtained by a position change matrix T, which is composed of a rotation R and a displacement t: T = [R t]; Then the depth d of the object to be photographed relative to the two plates is 1 , d 2 Satisfies the following formula: Among them, p 1 ,p 2 is a homogeneous coordinate with a coordinate of [x, y, 1], where x and y are the coordinates on the plate. Solving this gives the depth d 1 , d 2 .

7. A measuring device based on multi-view slit full-length X-ray photography, realizing the measuring method based on multi-view slit full-length X-ray photography as claimed in any one of claims 1 to 6, It is characterized in that The device comprises a plurality of longitudinal moving axes, a main control module, and a plurality of shooting components arranged on the longitudinal moving axes; each group of the shooting components comprises a radiation source and a flat plate, and the radiation source and the flat plate in the same group are arranged on different longitudinal moving axes, and are driven by the longitudinal moving axes to realize synchronous movement, and shoot the object to be shot to obtain a group of image pictures; The main control module is configured to receive a number of images taken by all the shooting components, and stitch the images together, compare the plate height of the current shooting component with the thickness of the parallel light emitted by the radiation source and constrained by the beam splitter; if the plate height is less than the parallel light thickness, stitch the adjacent image pictures according to the physical positions of the corresponding plates to obtain a stitched image; if the plate height is greater than the parallel light thickness, match the adjacent image pictures to obtain the best lateral offset and the best longitudinal offset between the adjacent image pictures, and stitch all the image pictures together according to the best lateral offset and the best longitudinal offset to obtain a stitched image; after all the image pictures corresponding to all the shooting components are stitched together, map all the ordinates and abscissas of all the stitched images to real physical position pixels.

8. A computer device, It is characterized in that It comprises a memory and a processor, wherein the memory stores computer-readable instructions, and when the computer-readable instructions are executed by the processor, the processor executes the measurement method based on multi-view slit full-length X-ray photography as described in any one of claims 1 to 6.

9. A storage medium storing computer-readable instructions, It is characterized in that When the computer-readable instructions are executed by one or more processors, the one or more processors execute the measurement method based on multi-view slit full-length X-ray photography as described in any one of claims 1 to 6.

Images