A panoramic image and source image correspondence method, system and storage medium

Abstract

This application relates to the field of image processing technology, and in particular to a method, system, and storage medium for mapping panoramic images to source images. The method for mapping panoramic images to source images includes the following steps: performing panoramic stitching to obtain a panoramic image; calculating a projection mapping matrix from each source image to the panoramic image; projecting the label matrix of each source image onto the projection mapping matrix to obtain a sequence label matrix for the panoramic image; and responding to a click command on the panoramic image to obtain the sequence number of the source image corresponding to the clicked point. By using the above method, the mapping relationship between the panoramic image and the source image can be obtained. When it is necessary to locate a local area later, the target local area can be located by clicking on the panoramic image.

Description

Technical Field

[0001] This application relates to the field of image processing technology, and in particular to a method, system, and storage medium for corresponding panoramic images and source images. Background Technology

[0002] With the development of imaging technology, the public has increasingly higher demands for the visual quality of images. They always hope to see panoramic images with higher resolution and wider viewing angles; therefore, how to achieve panoramic image stitching has become particularly important.

[0003] There are various panoramic stitching technologies in the present technology. However, these technologies can only obtain the final panoramic image, but cannot obtain the correspondence between the panoramic image and the source image. This will cause inconvenience for subsequent local location positioning. Therefore, how to obtain the correspondence between the panoramic image and the source image has become an urgent technical problem to be solved. Summary of the Invention

[0004] In view of the above problems, this application provides a method for mapping panoramic images to source images, thereby solving the technical problem that existing technologies cannot directly obtain the mapping relationship between panoramic images and source images. The specific technical solution is as follows:

[0005] A method for mapping panoramic images to source images, comprising the following steps:

[0006] A panoramic image is obtained by stitching together panoramic images.

[0007] Calculate the projection mapping matrix from each source image to the panoramic image;

[0008] The projection mapping matrix is ​​used to project and map the label matrix of each source image to obtain the sequence label matrix of the panoramic image.

[0009] In response to a click command on the panoramic image, obtain the sequence number of the source image corresponding to the clicked point.

[0010] Furthermore, the step of "calculating the projection mapping matrix from each source image to the panoramic image" specifically includes the following steps:

[0011] Extract feature points from each source image, and extract feature points from the panoramic image;

[0012] Perform feature point matching to obtain the target matching point;

[0013] Calculate the projection mapping matrix from each source image to the panoramic image.

[0014] Furthermore, the step of "projecting and mapping the label matrix of each source image through the projection mapping matrix to obtain the sequence label matrix of the panoramic image" specifically includes the following steps:

[0015] The label matrix of each source image is projected onto the projection mapping matrix to obtain the label matrix of each source image in the panoramic image. The label matrix is ​​then multiplied by the corresponding source image number to obtain the index label matrix of each source image in the panoramic image.

[0016] Furthermore, the step of "responding to a panoramic image click command and obtaining the source image number corresponding to the clicked point" specifically includes the following steps:

[0017] In response to a click command on the panoramic image, the clicked point is located at the corresponding coordinates on the panoramic image.

[0018] The marker matrix of the source image corresponding to the clicked point is obtained based on the corresponding coordinate position;

[0019] The source image number corresponding to the clicked point is obtained from the marked matrix.

[0020] Furthermore, the step of "performing panoramic stitching to obtain a panoramic image" specifically includes the following steps:

[0021] Collect images of the target scene and store them in a preset file;

[0022] Read the image from the preset file;

[0023] The images are stitched together to obtain a panoramic image.

[0024] Furthermore, the step of "and storing the target scene image to a preset file" specifically includes the following steps:

[0025] The target scene images are stored in a preset file in a uniform format;

[0026] The step of "reading the image sequence from the preset file" further includes the following steps:

[0027] Batch reading is performed using the GLOB function.

[0028] Furthermore, the step of "performing panoramic stitching of the images to obtain a panoramic image" specifically includes the following steps:

[0029] Feature point detection is performed on the image, and the feature points of the image are matched using the nearest neighbor method. The confidence of the best match is saved, and the homography matrix of the matched feature points of the two images is saved.

[0030] Remove matches between images with low confidence and determine the stitched set of matching images using the disjoint-set data structure algorithm;

[0031] Camera parameters are estimated for all stitched images to obtain the rotation matrix, and then the rotation matrix is ​​further adjusted using the bundle averaging method.

[0032] Perform horizontal or vertical waveform correction;

[0033] Projection stitching from the original image to a specified panoramic image, illumination compensation, and multi-band fusion.

[0034] To address the aforementioned technical problems, a computer-readable storage medium is also provided, with the specific technical solution as follows:

[0035] A computer-readable storage medium having a computer program stored thereon, the program being executed by a processor of any step in the aforementioned method for mapping a panoramic image to a source image.

[0036] To address the aforementioned technical issues, a system for mapping panoramic images to source images is also provided. The specific technical solution is as follows:

[0037] A system for mapping panoramic images to source images, comprising: a server and a client;

[0038] The server is used to: perform panoramic stitching to obtain a panoramic image; calculate the projection mapping matrix from each source image to the panoramic image; and project the label matrix of each source image through the projection mapping matrix to obtain the sequence label matrix of the panoramic image.

[0039] The client is used to: send a panoramic image click command to the server;

[0040] The server is also used to: respond to the client's panoramic image click command and obtain the sequence number of the source image corresponding to the clicked point.

[0041] Furthermore, it also includes: camera devices;

[0042] The camera device is used to: capture a target image and send the target image to the server.

[0043] The beneficial effects of this invention are: a method for corresponding a panoramic image and a source image, comprising the steps of: performing panoramic stitching to obtain a panoramic image; calculating the projection mapping matrix from each source image to the panoramic image; projecting the label matrix of each source image onto the projection mapping matrix to obtain the sequence label matrix of the panoramic image; and responding to a panoramic image click command to obtain the sequence number of the source image corresponding to the clicked point. By using the above method, the correspondence between the panoramic image and the source image can be obtained. When it is necessary to locate a local area later, the target local area can be located by clicking on the panoramic image.

[0044] The above description of the invention is merely an overview of the technical solution of this application. In order to enable those skilled in the art to better understand the technical solution of this application and to implement it based on the description and drawings, and to make the above-mentioned objectives and other objectives, features and advantages of this application easier to understand, the following description is provided in conjunction with the specific embodiments and drawings of this application. Attached Figure Description

[0045] The accompanying drawings are only used to illustrate the principles, implementation methods, applications, features, and effects of specific embodiments of this application and other related content, and should not be considered as limitations on this application.

[0046] In the accompanying drawings of the instruction manual:

[0047] Figure 1 The flowchart of a method for mapping panoramic images to source images, as described in a specific implementation, is as follows: Figure 1 ;

[0048] Figure 2 The flowchart of a method for mapping panoramic images to source images, as described in a specific implementation, is as follows: Figure 2 ;

[0049] Figure 3 The flowchart of a method for mapping panoramic images to source images, as described in a specific implementation, is as follows: Figure 3 ;

[0050] Figure 4 A schematic diagram of a computer-readable storage medium according to a specific embodiment;

[0051] Figure 5 This is a schematic diagram of a system for corresponding panoramic images and source images, as described in a specific implementation.

[0052] The reference numerals used in the above figures are explained as follows:

[0053] 400. Computer-readable storage media,

[0054] 500. A system for matching panoramic images with source images.

[0055] 501, Server-side

[0056] 502, Client. Detailed Implementation

[0057] To illustrate the possible application scenarios, technical principles, implementable specific solutions, and achievable objectives and effects of this application in detail, the following description, in conjunction with the listed specific embodiments and accompanying drawings, provides a detailed explanation. The embodiments described herein are merely illustrative of the technical solutions of this application and are therefore intended to limit the scope of protection of this application.

[0058] In this document, the term "embodiment" means that a specific feature, structure, or characteristic described in connection with an embodiment may be included in at least one embodiment of this application. The term "embodiment" appearing in various places throughout the specification does not necessarily refer to the same embodiment, nor does it specifically limit its independence or connection with other embodiments. In principle, in this application, as long as there are no technical contradictions or conflicts, the technical features mentioned in each embodiment can be combined in any way to form corresponding implementable technical solutions.

[0059] Unless otherwise defined, the technical terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application pertains; the use of related terms herein is merely for the purpose of describing particular embodiments and is not intended to limit this application.

[0060] In the description of this application, the term "and / or" is used to describe the logical relationship between objects, indicating that three relationships can exist. For example, A and / or B means: A exists, B exists, and A and B exist simultaneously. Additionally, the character " / " in this document generally indicates that the preceding and following objects have an "or" logical relationship.

[0061] In this application, 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 actual quantity, hierarchy or order relationship between these entities or operations.

[0062] Unless otherwise specified, the use of terms such as “comprising,” “including,” “having,” or other similar expressions in this application is intended to cover non-exclusive inclusion, which does not exclude the presence of additional elements in a process, method, or product that includes the stated elements, such that a process, method, or product that includes a list of elements may include not only those defined elements but also other elements not expressly listed, or elements inherent to such a process, method, or product.

[0063] Similar to the understanding in the Examination Guidelines, in this application, expressions such as "greater than," "less than," and "exceeding" are understood to exclude the stated number; expressions such as "above," "below," and "within" are understood to include the stated number. Furthermore, in the description of the embodiments in this application, "multiple" means two or more (including two), and similar expressions related to "multiple" are also understood in this way, such as "multiple groups" and "multiple times," unless otherwise explicitly specified.

[0064] As mentioned in the background section above, the existing technology only mentions panoramic stitching technology, but it cannot obtain the correspondence between the panoramic image and the source image. This application provides a new method for corresponding panoramic images and source images, so that no matter where the user clicks on any position in the panoramic image, the corresponding source image number can be obtained.

[0065] See below Figures 1 to 3 The technical solution for a method of mapping panoramic images to source images is described in detail below:

[0066] Figure 1 The flowchart illustrates a method for mapping panoramic images to source images. Figure 1 It includes steps S101 to S104.

[0067] In step S101, panoramic stitching is performed to obtain a panoramic image. For example... Figure 3 As shown, step S101 further includes steps S301 to S303.

[0068] In step S301, target scene images are acquired and stored in a preset file. In this embodiment, target scene images are acquired using a camera. Preferably, the camera is suspended at any fixed position within the scene, providing an overview of the entire scene. Rotating the camera allows for the acquisition of a series of scene images, which are then stored in a designated folder. The camera can be in a fixed position or moved to acquire images.

[0069] The step of "and storing the target scene image to a preset file" further includes the following steps: storing the target scene image to a preset file in a uniform format, such as bmp format.

[0070] In step S302, images are read from the preset file. Specifically, this can be done by batch reading using the GLOB function. These images belong to the same scene and were captured at the same time.

[0071] In step S303, the images are stitched together to obtain a panoramic image. This further includes the following steps:

[0072] The system performs SURF or ORB feature point detection on the images, matches feature points using the nearest neighbor method, saves the confidence score of the best match, and simultaneously saves the homography matrix of the matched feature points between the two images; deletes matches between images with low confidence scores, and determines the stitching set of matched images using a disjoint-set data structure algorithm; roughly estimates camera parameters for all stitched images, then calculates the rotation matrix, and further estimates the rotation matrix more accurately using the bundle averaging method; performs horizontal or vertical waveform correction; and performs projection stitching from the original image to a specified panoramic image, illumination compensation, and multi-band fusion.

[0073] After obtaining the panoramic image, in step S102, the projection mapping matrix from each source image to the panoramic image is calculated. Specifically, this includes the following steps: extracting feature points from each source image; extracting feature points from the panoramic image; performing feature point matching to obtain target matching points; and calculating the projection mapping matrix from each source image to the panoramic image.

[0074] In this embodiment, feature points are extracted for each source image using the SURF or ORB algorithm, and feature points are extracted for the panoramic image using the SURF or ORB algorithm. The nearest neighbor method is used to match the feature points between the two. Lowe's algorithm is used to obtain excellent matching points (i.e., target matching points, in order to exclude key points without matching relationship caused by image occlusion and background clutter). Then, the excellent matching points are used as parameters of the findHomography function to calculate the projection mapping matrix H from each source image to the panoramic image.

[0075] In step S103, the label matrix of each source image is projected and mapped using the projection mapping matrix to obtain the sequence label matrix of the panoramic image. Specifically, it also includes the following steps:

[0076] The label matrix of each source image is projected onto the projection mapping matrix H to obtain the label matrix of each source image in the panoramic image. The label matrix is ​​then multiplied by the corresponding source image number to obtain the index label matrix of each source image in the panoramic image. The label matrix of each source image is a matrix of the same size as the source image, with all values ​​being 1.

[0077] Specifically, the process can be as follows: Apply the warpPerspective function to the label matrix of each source image to perform a projection mapping transformation, thereby obtaining the label matrix of each source image in the panoramic image. This label matrix contains only 0 and 1 elements. Then multiply it by the corresponding source image number to obtain the index label matrix of each source image in the panoramic image. This index label matrix is ​​0 in the part without scene and is the index from 1 to the total number of source images n in the part with scene.

[0078] For example: if the source image labeling matrix is ​​A, the source image labeling matrix is ​​a matrix of the same size as the source image with all values ​​being 1, the panoramic image number labeling matrix is ​​B, the projection matrix is ​​H, and the source image number is N, where N = 1, 2, ..., then B = A * H * N.

[0079] In step S104, in response to the panoramic image click command, the sequence number of the source image corresponding to the clicked point is obtained. For example... Figure 2 As shown, step S104 further includes steps S201 to S203.

[0080] In step S201, in response to a panoramic image click command, the corresponding coordinates of the clicked point in the panoramic image are obtained. The user can click on any point in the panoramic image on the client side, and the corresponding coordinates of that point in the panoramic image are obtained through the click action.

[0081] In step S202, the marker matrix of the source image corresponding to the clicked point is obtained according to the corresponding coordinate position.

[0082] In step S203, the index of the source image corresponding to the clicked point is obtained according to the marker matrix. If the point is located in the overlapping part of multiple source images, multiple index results will appear. The one with the smallest index can be selected as the final result, or the corresponding multiple source image indices can be output. The specific method can be determined according to the actual situation.

[0083] A method for mapping panoramic images to source images includes the following steps: performing panoramic stitching to obtain a panoramic image; calculating a projection mapping matrix from each source image to the panoramic image; projecting the label matrix of each source image onto the projection mapping matrix to obtain a sequence label matrix for the panoramic image; and responding to a click command on the panoramic image to obtain the sequence number of the source image corresponding to the clicked point. This method allows for the acquisition of the mapping relationship between panoramic and source images. When it is necessary to locate a local area, clicking on the panoramic image will locate the target local area.

[0084] See below Figure 4 The following provides a detailed description of a computer-readable storage medium 400:

[0085] A computer-readable storage medium 400 has a computer program stored thereon, the program being executed by a processor of the following steps:

[0086] Step S101: Perform panoramic stitching to obtain a panoramic image. For example... Figure 3 As shown, step S101 further includes steps S301 to S303.

[0087] Step S301: Acquire target scene images and store them in a preset file. In this embodiment, target scene images are acquired using a camera. Preferably, the camera is suspended at any fixed position within the scene, providing an overview of the entire scene. Rotating the camera allows for the acquisition of a series of scene images, which are then stored in a designated folder. The camera can be in a fixed position or moved to acquire images.

[0088] The step of "and storing the target scene image to a preset file" further includes the following steps: storing the target scene image to a preset file in a uniform format, such as bmp format.

[0089] Step S302: Read images from the preset file. Specifically, this can be done by batch reading using the GLOB function. These images belong to the same scene and were captured at the same time.

[0090] Step S303 involves performing panoramic stitching on the images to obtain a panoramic image. This further includes the following steps:

[0091] The system performs SURF or ORB feature point detection on the images, matches feature points using the nearest neighbor method, saves the confidence score of the best match, and simultaneously saves the homography matrix of the matched feature points between the two images; deletes matches between images with low confidence scores, and determines the stitching set of matched images using a disjoint-set data structure algorithm; roughly estimates camera parameters for all stitched images, then calculates the rotation matrix, and further estimates the rotation matrix more accurately using the bundle averaging method; performs horizontal or vertical waveform correction; and performs projection stitching from the original image to a specified panoramic image, illumination compensation, and multi-band fusion.

[0092] After obtaining the panoramic image, in step S102, the projection mapping matrix from each source image to the panoramic image is calculated. Specifically, this includes the following steps: extracting feature points from each source image; extracting feature points from the panoramic image; performing feature point matching to obtain target matching points; and calculating the projection mapping matrix from each source image to the panoramic image.

[0093] In this embodiment, feature points are extracted for each source image using the SURF or ORB algorithm, and feature points are extracted for the panoramic image using the SURF or ORB algorithm. The nearest neighbor method is used to match the feature points between the two. Lowe's algorithm is used to obtain excellent matching points (i.e., target matching points, in order to exclude key points without matching relationship caused by image occlusion and background clutter). Then, the findHomography function is used to calculate the projection mapping matrix H from each source image to the panoramic image.

[0094] In step S103, the label matrix of each source image is projected and mapped using the projection mapping matrix to obtain the sequence label matrix of the panoramic image. Specifically, it also includes the following steps:

[0095] The label matrix of each source image is projected onto the projection mapping matrix to obtain the label matrix of each source image in the panoramic image. The label matrix is ​​then multiplied by the corresponding source image index to obtain the index label matrix of each source image in the panoramic image. The label matrix of each source image is a matrix of the same size as the source image, with all values ​​being 1.

[0096] Specifically, the process can be as follows: Apply the warpPerspective function to the label matrix of each source image to perform a projection mapping transformation, thereby obtaining the label matrix of each source image in the panoramic image. This label matrix contains only 0 and 1 elements. Then multiply it by the corresponding source image number to obtain the index label matrix of each source image in the panoramic image. This index label matrix is ​​0 in the part without scene and is the index from 1 to the total number of source images n in the part with scene.

[0097] For example: if the source image labeling matrix is ​​A, the source image labeling matrix is ​​a matrix of the same size as the source image with all values ​​being 1, the panoramic image number labeling matrix is ​​B, the projection matrix is ​​H, and the source image number is N, where N = 1, 2, ..., then B = A * H * N.

[0098] In step S104, in response to the panoramic image click command, the sequence number of the source image corresponding to the clicked point is obtained. For example... Figure 2 As shown, step S104 further includes steps S201 to S203.

[0099] In step S201, in response to a panoramic image click command, the corresponding coordinates of the clicked point in the panoramic image are obtained. The user can click on any point in the panoramic image on the client side, and the corresponding coordinates of that point in the panoramic image are obtained through the click action.

[0100] In step S202, the marker matrix of the source image corresponding to the clicked point is obtained according to the corresponding coordinate position.

[0101] In step S203, the index of the source image corresponding to the clicked point is obtained according to the marker matrix. If the point is located in the overlapping part of multiple source images, multiple index results will appear. The one with the smallest index can be selected as the final result, or the corresponding multiple source image indices can be output. The specific method can be determined according to the actual situation.

[0102] By performing the above steps, the computer-readable storage medium 400 can obtain the correspondence between the panoramic image and the source image. When it is necessary to locate a local area in the future, the target local area can be located by clicking on the panoramic image.

[0103] See below Figure 5 The following is a detailed explanation of a panoramic image-source image correspondence system 500:

[0104] A panoramic image and source image correspondence system 500 includes: a server 501 and a client 502; the server 501 is used to: perform panoramic stitching to obtain a panoramic image; calculate the projection mapping matrix from each source image to the panoramic image; and perform projection mapping on the label matrix of each source image through the projection mapping matrix to obtain the sequence label matrix of the panoramic image.

[0105] The client 502 is used to: send a panoramic image click command to the server 501;

[0106] The server 501 is also used to: respond to the panoramic image click command from the client 502 and obtain the sequence number of the source image corresponding to the clicked point.

[0107] Furthermore, it also includes: camera devices;

[0108] The camera device is used to: capture target images and send the target images to the server 501. Preferably, the camera device is suspended at a fixed position within the scene, providing an overview of the entire scene. By rotating the camera, a series of scene images can be captured and stored in a designated folder. The camera can be in a fixed position or moved to capture images. The target scene images are stored in a uniform format in a preset file, such as a BMP file.

[0109] The server 501 is further configured to: perform SURF or ORB feature point detection on the image, match the feature points of the image using the nearest neighbor method, save the confidence of the optimal match, and simultaneously save the homography matrix of the matching feature points of the two images; delete the matches between images with low confidence, and determine the stitching set of the matching images using the disjoint-set data structure algorithm; perform a rough estimation of camera parameters on all stitched image sets, then calculate the rotation matrix, and further estimate the rotation matrix more accurately using the bundle averaging method; perform horizontal or vertical waveform correction; and perform projection stitching from the original image to the specified panoramic image, illumination compensation, and multi-band fusion.

[0110] After obtaining the panoramic image, feature points are extracted from each source image and the panoramic image; feature point matching is performed to obtain target matching points; and the projection mapping matrix from each source image to the panoramic image is calculated.

[0111] In this embodiment, feature points are extracted for each source image using the SURF or ORB algorithm, and feature points are extracted for the panoramic image using the SURF or ORB algorithm. The nearest neighbor method is used to match the feature points between the two. Lowe's algorithm is used to obtain excellent matching points (i.e., target matching points, in order to exclude key points without matching relationship caused by image occlusion and background clutter). Then, the findHomography function is used to calculate the projection mapping matrix H from each source image to the panoramic image.

[0112] The label matrix of each source image is projected onto the projection mapping matrix H to obtain the label matrix of each source image in the panoramic image. The label matrix is ​​then multiplied by the corresponding source image number to obtain the index label matrix of each source image in the panoramic image. The label matrix of each source image is a matrix of the same size as the source image, with all values ​​being 1.

[0113] Specifically, the process can be as follows: Apply the warpPerspective function to the label matrix of each source image to perform a projection mapping transformation, thereby obtaining the label matrix of each source image in the panoramic image. This label matrix contains only 0 and 1 elements. Then multiply it by the corresponding source image number to obtain the index label matrix of each source image in the panoramic image. This index label matrix is ​​0 in the part without scene and is the index from 1 to the total number of source images n in the part with scene.

[0114] For example: if the source image labeling matrix is ​​A, the source image labeling matrix is ​​a matrix of the same size as the source image with all values ​​being 1, the panoramic image number labeling matrix is ​​B, the projection matrix is ​​H, and the source image number is N, where N = 1, 2, ..., then B = A * H * N.

[0115] In response to client 502's command to click on any point in the panoramic image, the system obtains the corresponding coordinates of that point within the panoramic image. Based on these coordinates, it calculates the marker matrix of the source image corresponding to the clicked point. Then, it calculates the sequence number of the source image corresponding to the clicked point based on the marker matrix. If the point is located in an overlapping area of ​​multiple source images, multiple sequence numbers will appear. The system can choose the smallest sequence number as the final result, or it can output the sequence numbers of all multiple source images, depending on the specific requirements.

[0116] By using the above system 500, the correspondence between the panoramic image and the source image can be obtained. When it is necessary to locate a local area later, the target local area can be located by clicking on the panoramic image.

[0117] Finally, it should be noted that although the above embodiments have been described in the text and drawings of this application, this should not limit the scope of patent protection of this application. Any technical solutions that are based on the essential concept of this application and utilize the content described in the text and drawings of this application, resulting in equivalent structural or procedural substitutions or modifications, as well as the direct or indirect application of the technical solutions of the above embodiments to other related technical fields, are all included within the scope of patent protection of this application.

Claims

1. A method for corresponding a panoramic image with a source image, characterized in that, Including the following steps: A panoramic image is obtained by stitching together panoramic images. Calculate the projection mapping matrix from each source image to the panoramic image; The projection mapping matrix is ​​used to project and map the label matrix of each source image to obtain the sequence label matrix of the panoramic image. In response to a click command on the panoramic image, obtain the sequence number of the source image corresponding to the clicked point. The step of "calculating the projection mapping matrix from each source image to the panoramic image" further includes the following steps: Extract feature points from each source image, and extract feature points from the panoramic image; Perform feature point matching to obtain the target matching point; Calculate the projection mapping matrix from each source image to the panoramic image; The step of "performing panoramic stitching to obtain a panoramic image" further includes the following steps: Collect images of the target scene and store them in a preset file; Read the image from the preset file; The panoramic image is obtained by stitching the images together, and the process further includes the following steps: Feature point detection is performed on the image, and the feature points of the image are matched using the nearest neighbor method. The confidence of the best match is saved, and the homography matrix of the matched feature points of the two images is saved. Remove matches between images with low confidence and determine the stitched set of matching images using the disjoint-set data structure algorithm; Camera parameters are estimated for all stitched images to obtain the rotation matrix, and then the rotation matrix is ​​further adjusted using the bundle averaging method. Perform horizontal or vertical waveform correction; Projection stitching from the original image to a specified panoramic image, illumination compensation and multi-band fusion; The sequence number marking matrix of the panoramic image is labeled B, where B = A * H * N, and N is the source image sequence number; A is the source image marking matrix; and H is the projection mapping matrix.

2. The method for corresponding a panoramic image and a source image according to claim 1, characterized in that, The step of "responding to a panoramic image click command and obtaining the source image number corresponding to the clicked point" further includes the following steps: In response to a click command on the panoramic image, the clicked point is located at the corresponding coordinates on the panoramic image. The marker matrix of the source image corresponding to the clicked point is obtained based on the corresponding coordinate position; The source image number corresponding to the clicked point is obtained from the marked matrix.

3. The method for corresponding a panoramic image and a source image according to claim 1, characterized in that, The step of "and storing the target scene image to a preset file" further includes the following steps: The target scene images are stored in a preset file in a uniform format; The step of "reading the image sequence from the preset file" further includes the following steps: Batch reading is performed using the GLOB function.

4. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the program is executed by the processor, it implements the steps as described in any one of claims 1 to 3.

5. A system for mapping panoramic images to source images, characterized in that, include: Server and client; The server is used to: perform panoramic stitching to obtain a panoramic image; and calculate the projection mapping matrix from each source image to the panoramic image. The projection mapping matrix is ​​used to project and map the label matrix of each source image to obtain the sequence label matrix of the panoramic image; the sequence label matrix of the panoramic image is labeled as B, B=A*H*N, where N is the source image sequence number; A is the source image label matrix; H is the projection mapping matrix; The client is used to: send a panoramic image click command to the server; The server is also used to: respond to the client's panoramic image click command and obtain the sequence number of the source image corresponding to the clicked point; The step of "calculating the projection mapping matrix from each source image to the panoramic image" further includes the following steps: Extract feature points from each source image, and extract feature points from the panoramic image; Perform feature point matching to obtain the target matching point; Calculate the projection mapping matrix from each source image to the panoramic image; The step of "performing panoramic stitching to obtain a panoramic image" further includes the following steps: Collect images of the target scene and store them in a preset file; Read the image from the preset file; The panoramic image is obtained by stitching the images together, and the process further includes the following steps: Feature point detection is performed on the image, and the feature points of the image are matched using the nearest neighbor method. The confidence of the best match is saved, and the homography matrix of the matched feature points of the two images is saved. Remove matches between images with low confidence and determine the stitched set of matching images using the disjoint-set data structure algorithm; Camera parameters are estimated for all stitched images to obtain the rotation matrix, and then the rotation matrix is ​​further adjusted using the bundle averaging method. Perform horizontal or vertical waveform correction; Projection stitching from the original image to a specified panoramic image, illumination compensation, and multi-band fusion.

6. The panoramic image and source image correspondence system according to claim 5, characterized in that, It also includes: camera devices; The camera device is used to: capture a target image and send the target image to the server.

Images