Aircraft-oriented multi-view image adaptive splicing and perception method and system

Abstract

The application discloses a kind of multi-view image adaptive splicing and perception method and system for aircraft. Including: obtaining multi-view image and extracting feature points, obtaining reliable matching point pairs by eliminating false matching. Grid area is divided for non-planar scene, and local homography model is constructed. On this basis, a multi-constraint optimization model is constructed, and each local homography matrix is optimized. At the same time, the brightness difference of the overlapping area is calculated, and the space brightness adjustment model is constructed to obtain the optimal parameters of photometric consistency. Finally, the geometric and photometric optimization parameters are fused, and the splicing image is output. The application effectively improves the splicing accuracy and visual consistency of multi-view image of aircraft by adaptive processing of non-planar scene with local homography.

Description

Technical Field

[0001] This invention belongs to the field of environmental perception and computer vision technology, and in particular relates to a multi-view image adaptive stitching and perception method and system for aircraft. Background Technology

[0002] With the rapid development of the low-altitude economy, new types of manned aircraft, represented by eVTOL (electric vertical takeoff and landing aircraft), are gradually entering application scenarios such as urban air traffic, emergency rescue, and urban logistics. Compared with traditional general aviation aircraft, eVTOL has the advantages of flexible takeoff and landing, low operating costs, and strong adaptability to complex urban environments, making it an important component of the future three-dimensional urban transportation system.

[0003] In practical applications, eVTOL manned aircraft often need to perform flight missions in densely packed, structurally complex near-wall environments, such as navigating between urban buildings, hovering and observing near walls, window rescues, and navigation in confined spaces. These scenarios are characterized by small spatial scale, severe obstruction, drastic changes in lighting, and strong dynamic interference, which places extremely high demands on the aircraft's environmental perception capabilities.

[0004] In multi-view vision systems, due to the drastic attitude changes, significant vibration effects, and large differences in viewing angles during flight, traditional methods based on global homography or static stitching models are difficult to adapt to complex non-planar structural environments. This can easily lead to problems such as stitching misalignment, geometric distortion, and inconsistent light intensity, thereby affecting the reliability of overall environmental perception.

[0005] In summary, there is an urgent need to propose a multi-view robust perception and image stitching method for low-altitude economic eVTOL manned aircraft. Summary of the Invention

[0006] The purpose of this invention is to provide a multi-view image adaptive stitching and perception method and system for aircraft. By constructing an image stitching model that integrates geometric constraints and motion compensation, and combining it with a photometric adaptive adjustment method based on the consistency of overlapping regions, high-precision fusion and stable output of multi-view images can be achieved.

[0007] To achieve the objectives of this invention, in one aspect, this invention provides a multi-view image adaptive stitching and perception method for aircraft, comprising the following steps:

[0008] S1. Images are acquired through multi-view cameras mounted on the aircraft, and preprocessed and feature extracted to obtain a set of feature points; then feature points are matched between adjacent view images, and mismatched points are removed to obtain a set of reliable matching point pairs.

[0009] S2. For non-planar scenes, the image is divided into several local regions based on a grid; within each local region, a local homography model is constructed by estimating the local geometric transformation relationship based on the set of matching point pairs.

[0010] S3. Based on the local homography model, a multi-constraint optimization model is constructed to form a joint optimization objective function. The free parameters of each local homography matrix are optimized and solved to obtain the optimal homography matrix set.

[0011] S4. The splicing area between the adjacent viewpoint images is calculated and a spatial brightness adjustment model is constructed by the pixel brightness difference in the overlapping area to obtain the photometric consistency optimization model, and its optimal parameters are output.

[0012] S5. The optimal parameters of the multi-constraint optimization model and the photometric consistency optimization model are fused together to output the image stitching result.

[0013] S1 is a set of reliable matching points filtered by a random sampling consensus algorithm combined with reprojection error constraints to eliminate false matches, as shown in the following formula:

[0014] ;

[0015] in, It is the set of reliable matching points after filtering. It is the homography matrix estimated by the random sampling consensus algorithm. It is a homogeneous coordinate normalization function. It is the first Images captured by a camera and the Images captured by a camera Matching feature points, It is the reprojection error threshold.

[0016] The local homography model of S2 is shown in the following equation:

[0017] ;

[0018] in, It is the first The homography matrix of each region yes Real matrix space of order;

[0019] The transformation relationship must be satisfied:

[0020] ;

[0021] in, These are the homogeneous coordinates in the aligned image. These are the homogeneous coordinates in the input image.

[0022] The multi-constraint optimization model of S3 includes feature alignment, region continuity, and boundary consistency constraints.

[0023] Among them, regional continuity constraints The specific formula is as follows:

[0024] ;

[0025] in, Represents a set of adjacent regions. It is the F-norm. Is with the first The adjacent regions of the first The homography matrix of a local region These are adjacent region index pairs;

[0026] A multi-constraint optimization model is constructed based on the above constraints. :

[0027] ;

[0028] in, It is the weight of the regional continuity constraint. It is the boundary consistency constraint weight. It is feature alignment error. It is a boundary consistency constraint;

[0029] By minimizing the objective function For each local region homography matrix The optimal matrix is ​​obtained by jointly optimizing all free parameters. .

[0030] The photometric consistency optimization model of S4 is described, and its optimal parameters are output as follows:

[0031] The photometric consistency optimization model The specific formula is as follows:

[0032] ;

[0033] in, It is the first and the The set of pixels in the overlapping region of an image. It refers to the pixel position within the overlapping area. It is the energy of brightness difference. It is an image In position pixel values, It is a spatial brightness adjustment model;

[0034] Then, by minimizing the brightness difference, its optimal parameters are solved. To achieve a smooth transition between spliced ​​areas, among which, It is brightness gain. It's a brightness offset.

[0035] The fusion model of S5 is specifically shown in the following formula:

[0036] ;

[0037] in, It is an image fusion function. This is the fused output image. It is the optimized set of local geometric transformation parameters. It is the set of all input images after photometric compensation;

[0038] Images after geometric transformation and photometric compensation are fused using a weighted average method:

[0039] ;

[0040] in, Is the input image in The weight of the position, It is the input image exist The pixel value after local geometric transformation and photometric compensation. Is the output image in The pixel value at that location.

[0041] On the other hand, the present invention also provides a system for adaptive stitching and perception of multi-view images for aircraft, comprising the following modules:

[0042] The image acquisition and feature extraction module is used to control the multi-view camera on the aircraft to acquire images, preprocess the images, and extract feature points from the images from each view.

[0043] The feature matching and mismatch removal module is used to match feature points between adjacent viewpoint images and remove mismatch points through geometric consistency constraints to obtain a reliable set of matching point pairs.

[0044] The local homography modeling module is used for non-planar scenes to divide the image into grid-like local regions, estimate the local geometric transformation relationship in each local region based on the matching point pairs, and construct a local homography model.

[0045] The multi-constraint optimization solution module is used to construct a joint optimization objective function containing multiple geometric constraints based on the local homography model, optimize the free parameters of each local homography matrix, and obtain the optimal homography matrix set.

[0046] The photometric consistency optimization module is used to analyze the pixel brightness differences in the overlapping areas of the stitched areas of adjacent viewpoint images, construct a spatial brightness adjustment model, optimize photometric consistency parameters, and eliminate stitching seams and brightness inconsistencies.

[0047] The fusion and stitching output module is used to fuse the optimal parameters of geometry and photometry to complete image stitching and output the final stitched image.

[0048] An electronic device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the multi-view image adaptive stitching and perception method as described above.

[0049] A non-transitory computer-readable storage medium stores computer instructions for causing the computer to execute the above-described multi-view image adaptive stitching and perception method.

[0050] A computer program product includes computer program instructions that, when run on a computer, cause the computer to execute the aforementioned multi-view image adaptive stitching and perception method.

[0051] The significant advancement of this invention compared to existing technologies lies in:

[0052] (1) This invention divides the image into grid regions and estimates the geometric transformation relationship by using a local homography model. This can effectively deal with the parallax changes in non-planar scenes, significantly improve the image alignment accuracy in complex scenes, and avoid the ghosting and misalignment phenomena caused by traditional global homography methods.

[0053] (2) The present invention constructs a multi-constraint joint optimization objective function that includes feature alignment constraints, regional continuity constraints and boundary consistency constraints. It can coordinately optimize the free parameters of each local homography matrix, ensure smooth transition of geometric transformation between adjacent regions, and enhance the structural integrity and spatial consistency of the splicing result.

[0054] (3) The present invention introduces a spatial brightness adjustment model based on the pixel brightness difference calculation in the overlapping area. By solving the optimal brightness gain and offset parameters, it effectively eliminates the splicing seam and brightness change caused by the change of illumination or the difference of camera response, and realizes the smooth transition and light consistency of the splicing area.

[0055] (4) The present invention integrates the optimal parameters of the geometric optimization model and the photometric optimization model, and uses a weighted average method to fuse the multi-view images after local homography transformation and photometric compensation. This can maintain the geometric alignment accuracy between images and ensure the visual color and brightness continuity, and finally output a high-quality panoramic stitched image.

[0056] (5) The present invention uses a random sampling consensus algorithm combined with reprojection error constraints to remove mismatched points, which can effectively screen out unreliable feature matching pairs, improve the robustness and accuracy of local homography model estimation, and provide high-quality input data for subsequent optimization;

[0057] (6) This invention is specifically designed for the application scenarios of aircraft equipped with multi-view cameras. It can adapt to complex environments such as large changes in viewing angle, uneven distribution of scene depth, and variable lighting conditions during aerial photography. It has good self-adaptability and engineering practical value.

[0058] To more clearly illustrate the functional characteristics and structural parameters of the present invention, further explanation is provided below in conjunction with the accompanying drawings and specific embodiments. Attached Figure Description

[0059] The accompanying drawings, which are included to provide a further understanding of the invention and form part of this application, illustrate exemplary embodiments of the invention and, together with their description, serve to explain the invention and do not constitute an undue limitation thereof. In the drawings:

[0060] Figure 1 This is a flowchart of the steps of the present invention;

[0061] Figure 2 This is a schematic diagram illustrating the splicing process and results of the present invention; Figure 2 In the middle (a), two images to be stitched are input. Figure 2 (b) is a schematic diagram of the consistency residuals of the overlapping region boundaries after optimization. Figure 2 Image (c) is the stitched image. Detailed Implementation

[0062] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0063] This invention provides a multi-view image adaptive stitching and perception method for aircraft, combining... Figure 1 This includes the following steps:

[0064] S1. Acquisition of multi-view images and motion information: Images are acquired through multi-view cameras mounted on the aircraft, and preprocessed and feature extracted to obtain a set of feature points; then feature points are matched between adjacent view images, and mismatched points are removed to obtain a set of reliable matching point pairs.

[0065] S2. Local Region Segmentation and Geometric Modeling: For complex non-planar scenes, the image is divided into several local regions based on a grid; within each local region, a local homography model is constructed by estimating the local geometric transformation relationship based on the set of matching point pairs.

[0066] S3. Multi-constraint splicing optimization modeling: Based on the local homography model, a multi-constraint optimization model is constructed to form a joint optimization objective function. The free parameters of each local homography matrix are optimized and solved to obtain the optimal homography matrix set.

[0067] S4. Adaptive image enhancement based on overlapping region consistency: The splicing region between the adjacent viewpoint images is calculated and a spatially related brightness adjustment model is constructed based on the pixel brightness difference in the overlapping region to obtain a photometric consistency optimization model, and its optimal parameters are output.

[0068] S5. Generation and Perception Output of Stitching Results: The multi-constraint optimization model and the photometric consistency optimization model are fused together to output the image stitching results.

[0069] S1 specifically includes the following steps:

[0070] S1-1, Acquiring image data via a multi-view camera mounted on the aircraft. : ;in, It refers to the number of cameras. It is the first Images captured by a camera, It's the camera serial number. This is the total number of cameras;

[0071] S1-2. Preprocess and extract features from the image data to obtain a set of feature points. : ;in, The first one is the one mentioned Images captured by a camera The Middle Coordinates of feature points It is a feature point index. It is an image The Middle The x-coordinates of each feature point It is an image The Middle The ordinates of each feature point;

[0072] S1-3. Based on the set of feature points, obtain a set of matching point pairs by performing feature matching between images from adjacent viewpoints. : ;in, It is the first Images captured by a camera and the Images captured by a camera Matching feature points;

[0073] S1-4. The matching feature points are filtered using the Random Sample Consensus (RANSAC) algorithm to remove mismatched points and obtain a reliable set.

[0074] The random sampling consensus algorithm in S1-4, combined with reprojection error constraints, removes false matches. The resulting set of reliable matching points is shown in the following formula:

[0075] ;

[0076] in, It is the set of reliable matching points after filtering. It is the homography matrix estimated by the random sampling consensus algorithm. It is a homogeneous coordinate normalization function. It is the reprojection error threshold, in pixels, and is set to 2-5 pixels based on experimental experience.

[0077] S2 specifically includes the following steps:

[0078] S2-1. Divide the image into several local regions based on a grid: ;in, It is the first Images captured by a camera, It is the first A local area, This is the total number of regions;

[0079] S2-2. Within each local region, a local homography model is constructed by estimating the local geometric transformation relationship based on the matching point pairs.

[0080] The specific formula is as follows:

[0081] ;

[0082] Satisfy the transformation relationship:

[0083] ;

[0084] in, It is the first The homography matrix of each region yes Real matrix space of order 1 These are the homogeneous coordinates in the aligned image. These are the homogeneous coordinates in the input image. , It is an image The x and y coordinates of the middle pixel , It is an image The x and y coordinates of the middle pixel It is a matrix transpose operation;

[0085] The first homography matrix of each region Expanded as:

[0086] ;

[0087] in, , It is the two-dimensional coordinates of the input image. , These are the transformed homogeneous coordinate components. It is the scale factor used for normalization.

[0088] Combination Figure 2 The multi-constraint optimization model of S3 includes feature alignment, region continuity, and boundary consistency constraints, specifically:

[0089] The feature alignment error is shown in the following formula:

[0090] ;

[0091] in, It belongs to the region matching point, Homogeneous coordinate normalization function, Euclidean distance It is a set of matching feature points. It is feature alignment error;

[0092] The regional continuity constraint is shown in the following formula:

[0093] ;

[0094] in, Represents a set of adjacent regions. It is the F-norm. It is the first The homography matrix of a local region Is with the first The adjacent regions of the first The homography matrix of a local region These are adjacent region index pairs. It is a regional continuity constraint error;

[0095] Boundary consistency constraints are shown in the following formula:

[0096] ;

[0097] in, It is a region and The set of boundary pixels, This is the scripture Mapped pixel values, It is a region through After transformation Pixel value at that location, These are image coordinates. This is the boundary consistency error, a constraint used to ensure pixel continuity at the boundaries of adjacent regions and reduce stitching seams.

[0098] Based on the above constraints, a multi-constraint optimization model is constructed. :

[0099] ;

[0100] in, and These are the weighting coefficients. It is the weight of the regional continuity constraint. It is the boundary consistency constraint weight. and The settings are adaptively configured based on scene complexity and stitching effect. It is feature alignment error. It is a regional continuity error, a boundary consistency error;

[0101] By minimizing the objective function For each local region homography matrix The optimal matrix is ​​obtained by jointly optimizing all free parameters. ,in, It is the first The optimal homography matrix is ​​obtained by joint optimization of several local regions.

[0102] Specifically, S4 is:

[0103] S4-1. Extract the overlapping region between adjacent images. : ;

[0104] S4-2, Calculate the pixel brightness difference within the overlapping region. : ;in, It is an image In position pixel values, It is an image In position Pixel values;

[0105] S4-3. Construct a spatially relevant brightness adjustment model : ;in, It is a gain term. It is a bias term. These are the enhanced pixel values;

[0106] S4-4. Establish optimization goals : ;in, It is the first and the The set of pixels in the overlapping region of an image. It refers to the pixel position within the overlapping area. It is the energy of brightness difference;

[0107] S4-5, Solve by minimizing the brightness difference. To achieve a smooth transition between spliced ​​areas, among which, It is the optimal parameter for minimizing the brightness difference in the splicing area. It is brightness gain. It's a brightness offset.

[0108] The fusion model of S5 is specifically shown in the following formula:

[0109] ;

[0110] in, It is an image fusion function. This is the fused output image. It is the optimized set of local geometric transformation parameters. It is the set of all input images after photometric compensation. , These are the brightness adjustment parameters: gain and offset.

[0111] Images after geometric transformation and photometric compensation are fused using a weighted average method:

[0112] ;

[0113] in, Is the input image in The weight of the position, It is the input image exist The pixel value after local geometric transformation and photometric compensation. Is the output image in The pixel value at that location.

[0114] The present invention provides a system for adaptive stitching and sensing of multi-view images of aircraft, comprising the following modules:

[0115] The image acquisition and feature extraction module is used to control the multi-view camera on the aircraft to acquire images, preprocess the images, and extract feature points from the images from each view.

[0116] The feature matching and mismatch removal module is used to match feature points between adjacent viewpoint images and remove mismatch points through geometric consistency constraints to obtain a reliable set of matching point pairs.

[0117] The local homography modeling module is used for non-planar scenes to divide the image into grid-like local regions, estimate the local geometric transformation relationship in each local region based on the matching point pairs, and construct a local homography model.

[0118] The multi-constraint optimization solution module is used to construct a joint optimization objective function containing multiple geometric constraints based on the local homography model, optimize the free parameters of each local homography matrix, and obtain the optimal homography matrix set.

[0119] The photometric consistency optimization module is used to analyze the pixel brightness differences in the overlapping areas of the stitched areas of adjacent viewpoint images, construct a spatial brightness adjustment model, optimize photometric consistency parameters, and eliminate stitching seams and brightness inconsistencies.

[0120] The fusion and stitching output module is used to fuse the optimal parameters of geometry and photometry to complete image stitching and output the final stitched image.

[0121] like Figure 2 As shown, traverse the above steps, Figure 2 (a) is the input of two images to be stitched together. The two images were captured by different cameras and have a large overlapping area between their fields of view. The lighting conditions at the camera shooting angles are different. Figure 2 (b) is a schematic diagram of the boundary consistency residual of the overlapping region after optimization. The black area represents the non-overlapping area or the area with no effective pixels, and the colored highlight texture represents the pixel difference that still exists between the two images in the overlapping area. By introducing the pixel continuity constraint of the adjacent region boundary, the generation of pixel abrupt changes in the overlapping area can be effectively reduced. Figure 2 Image (c) is the stitched image. After feature alignment, local region optimization, and image fusion processing, the overall transition of the stitched image is natural.

[0122] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0123] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A multi-view image adaptive stitching and perception method for aircraft, characterized in that, Includes the following steps: S1. Images are acquired through multi-view cameras mounted on the aircraft, and preprocessed and feature extracted to obtain a set of feature points; then feature points are matched between adjacent view images, and mismatched points are removed to obtain a set of reliable matching point pairs. S2. For non-planar scenes, the image is divided into several local regions based on a grid; within each local region, a local homography model is constructed by estimating the local geometric transformation relationship based on the set of matching point pairs. S3. Based on the local homography model, a multi-constraint optimization model is constructed to form a joint optimization objective function. The free parameters of each local homography matrix are optimized and solved to obtain the optimal homography matrix set. S4. The splicing area between the adjacent viewpoint images is calculated and a spatial brightness adjustment model is constructed by the pixel brightness difference in the overlapping area to obtain the photometric consistency optimization model, and its optimal parameters are output. S4-1. Extract the overlapping region between adjacent images. : ; S4-2, Calculate the pixel brightness difference within the overlapping region. : ;in, It is an image In position pixel values, It is an image In position Pixel values; S4-3. Construct a spatially relevant brightness adjustment model : ;in, It is a gain term. It is a bias term. These are the enhanced pixel values; S4-4. Establish a photometric consistency optimization model : ;in, It is the first and the The set of pixels in the overlapping region of an image. It refers to the pixel position within the overlapping area. It is the energy of brightness difference; S4-5, Solve by minimizing the brightness difference. To achieve a smooth transition between spliced ​​areas, among which, It is the optimal parameter for minimizing the brightness difference in the splicing area. It is brightness gain. It's a brightness shift; S5. The optimal parameters of the multi-constraint optimization model and the photometric consistency optimization model are fused together to output the image stitching result.

2. The multi-view image adaptive stitching and perception method for aircraft according to claim 1, characterized in that, S1 is a set of reliable matching points filtered by a random sampling consensus algorithm combined with reprojection error constraints to eliminate false matches, as shown in the following formula: ； in, It is the set of reliable matching points after filtering. It is the homography matrix estimated by the random sampling consensus algorithm. It is a homogeneous coordinate normalization function. It is the first Images captured by a camera and the Images captured by a camera Matching feature points, It is the reprojection error threshold.

3. The multi-view image adaptive stitching and perception method for aircraft according to claim 2, characterized in that, The local homography model of S2 is shown in the following equation: ； in, It is the first The homography matrix of each region yes Real matrix space of order; The transformation relationship must be satisfied: ； in, These are the homogeneous coordinates in the aligned image. These are the homogeneous coordinates in the input image.

4. The multi-view image adaptive stitching and perception method for aircraft according to claim 3, characterized in that, The multi-constraint optimization model of S3 includes feature alignment, region continuity, and boundary consistency constraints. Among them, regional continuity constraints The specific formula is as follows: ； in, Represents a set of adjacent regions. It is the F-norm. Is with the first The adjacent regions of the first The homography matrix of a local region These are adjacent region index pairs; A multi-constraint optimization model is constructed based on the above constraints. : ； in, It is the weight of the regional continuity constraint. It is the boundary consistency constraint weight. It is feature alignment error. It is a boundary consistency constraint; By minimizing the objective function For each local region homography matrix The optimal matrix is ​​obtained by jointly optimizing all free parameters. .

5. The multi-view image adaptive stitching and perception method for aircraft according to claim 4, characterized in that, The fusion model of S5 is specifically shown in the following formula: ； in, It is an image fusion function. This is the fused output image. It is the optimized set of local geometric transformation parameters. It is the set of all input images after photometric compensation; Images after geometric transformation and photometric compensation are fused using a weighted average method: ； in, Is the input image in The weight of the position, It is the input image exist The pixel value after local geometric transformation and photometric compensation. Is the output image in The pixel value at that location.

6. A system for implementing the multi-view image adaptive stitching and sensing method for aircraft as described in any one of claims 1-5, characterized in that, Includes the following modules: The image acquisition and feature extraction module is used to control the multi-view camera on the aircraft to acquire images, preprocess the images, and extract feature points from the images from each view. The feature matching and mismatch removal module is used to match feature points between adjacent viewpoint images and remove mismatch points through geometric consistency constraints to obtain a reliable set of matching point pairs. The local homography modeling module is used for non-planar scenes to divide the image into grid-like local regions, estimate the local geometric transformation relationship in each local region based on the matching point pairs, and construct a local homography model. The multi-constraint optimization solution module is used to construct a joint optimization objective function containing multiple geometric constraints based on the local homography model, optimize the free parameters of each local homography matrix, and obtain the optimal homography matrix set. The photometric consistency optimization module is used to analyze the pixel brightness differences in the overlapping areas of the stitched areas of adjacent viewpoint images, construct a spatial brightness adjustment model, optimize photometric consistency parameters, and eliminate stitching seams and brightness inconsistencies. The fusion and stitching output module is used to fuse the optimal parameters of geometry and photometry to complete image stitching and output the final stitched image.

7. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the method as described in any one of claims 1 to 5.

8. A non-transitory computer-readable storage medium, characterized in that, The non-transitory computer-readable storage medium stores computer instructions for causing the computer to perform the method as described in any one of claims 1 to 5.

9. A computer program product comprising computer program instructions, characterized in that, When the computer program instructions are executed on a computer, the computer causes the computer to perform the method as described in any one of claims 1 to 5.

Images