Distributed picture distortion correction method and system, electronic device and storage medium

Abstract

The application provides a distributed picture distortion correction method, system, electronic equipment and storage medium, and the method comprises the following steps: obtaining a distortion calibration matrix of a target camera according to a chessboard picture; creating an RDD (Resilient Distributed Datasets) task according to a plurality of pictures to be corrected in an HDFS (Hadoop Distributed File System); correcting each picture to be corrected in the RDD task by using a foreach operator according to the distortion calibration matrix; and uploading the corrected picture to the HDFS until the correction of all the pictures to be corrected is completed. The application corrects each picture to be corrected in the RDD task by using the foreach operator according to the distortion calibration matrix to obtain a corrected picture, thereby combining the distortion calibration matrix with big data picture processing, achieving the purpose of simultaneously correcting a large number of distorted pictures in batches, and greatly improving the picture correction efficiency.

Description

Technical Field

[0001] This invention relates to the field of image correction technology, and more specifically, to a distributed image distortion correction method, system, electronic device, and storage medium. Background Technology

[0002] A fisheye camera is a camera equipped with a fisheye lens. The main advantage of a fisheye lens is its wide field of view, typically reaching 220° or 230°, which allows for close-up shooting of large areas of scenery. When shooting close to a subject, a fisheye lens creates a very strong perspective effect, emphasizing the contrast between objects appearing larger when closer and smaller when farther away. The principle is that the shorter the focal length, the wider the field of view, but the stronger the distortion caused by optical principles. As a result, apart from the object in the center of the frame remaining unchanged, other objects that should be horizontal or vertical are altered accordingly.

[0003] With the development of technology, in order to improve the wide angle and enhance the user experience, the lenses of vehicle cameras have been gradually replaced with fisheye lenses. As a result, distortion in the images captured by vehicle cameras has become a thorny issue that reduces the user experience. Therefore, how to quickly and efficiently process and correct image distortion is an urgent problem to be solved. Summary of the Invention

[0004] This invention addresses the technical problems existing in the prior art by providing a distributed image distortion correction method, system, electronic device, and storage medium to solve the problem of how to achieve fast and efficient image distortion correction.

[0005] According to a first aspect of the present invention, a distributed image distortion correction method is provided, comprising:

[0006] Obtain the distortion calibration matrix of the target camera from the chessboard image;

[0007] The task of creating an RDD resilient distributed dataset based on the distortion calibration matrix and multiple images to be corrected in the HDFS distributed file system;

[0008] The foreach operator is used to correct each image in the RDD Resilient Distributed Dataset task;

[0009] The corrected images are uploaded to the HDFS distributed file system until all images to be corrected are corrected.

[0010] Based on the above technical solution, the present invention can also be improved as follows.

[0011] Optionally, the step of obtaining the distortion calibration matrix of the target camera based on the chessboard image includes:

[0012] Obtain four chessboard images from different angles: front, back, left, and right.

[0013] The matrix mapping coefficients of the target camera are obtained using a preset visual processing algorithm and the chessboard image;

[0014] The distortion calibration matrix of the target camera is calculated based on the matrix mapping coefficients.

[0015] Optionally, the step of using the foreach operator to correct each image to be corrected in the RDD Resilient Distributed Dataset task according to the distortion calibration matrix includes:

[0016] The parameters of the preset visual processing algorithm are initialized according to the distortion calibration matrix;

[0017] The Spark computing engine is used to correct each image to be corrected in the RDD Resilient Distributed Dataset task according to the initialized visual processing algorithm.

[0018] Optionally, the preset visual processing algorithm is a visual processing algorithm based on OpenCV.

[0019] Optionally, after the step of calculating the distortion calibration matrix of the target camera based on the matrix mapping coefficients, the method further includes:

[0020] Obtain a preset free scaling factor, and optimize the distortion calibration matrix based on the preset free scaling factor.

[0021] Optionally, the step of obtaining a preset free scaling factor and optimizing the distortion calibration matrix based on the preset free scaling factor includes:

[0022] When the preset free scaling factor is 0, the distortion calibration matrix is ​​pruned and optimized.

[0023] Optionally, the step of obtaining a preset free scaling factor and optimizing the distortion calibration matrix based on the preset free scaling factor further includes:

[0024] When the preset free scaling factor is 1, the distortion calibration matrix is ​​optimized by adding features.

[0025] According to a second aspect of the present invention, a distributed image distortion correction system is provided, comprising:

[0026] The distortion calculation module is used to obtain the intrinsic parameter matrix of the target camera based on the chessboard image;

[0027] The task creation module is used to create an RDD elastic distributed dataset task from multiple images to be corrected in the HDFS distributed file system.

[0028] The image correction module is used to correct each image to be corrected in the RDD Resilient Distributed Dataset task using the foreach operator based on the distortion calibration matrix;

[0029] The image upload module is used to upload the corrected images to the HDFS distributed file system until all images to be corrected are corrected.

[0030] According to a third aspect of the present invention, an electronic device is provided, including a memory and a processor, wherein the processor is configured to execute a computer management program stored in the memory to implement the steps of any of the distributed image distortion correction methods described in the first aspect.

[0031] According to a fourth aspect of the present invention, a computer-readable storage medium is provided, on which a computer management class program is stored, wherein when executed by a processor, the computer management class program implements the steps of any of the distributed image distortion correction methods of the first aspect described above.

[0032] This invention provides a distributed image distortion correction method, system, electronic device, and storage medium. The method involves obtaining the distortion calibration matrix of the target camera based on a chessboard image; creating an RDD (Resilient Distributed Dataset) task from multiple images to be corrected in the HDFS distributed file system; using a foreach operator to correct each image in the RDD task based on the distortion calibration matrix; and uploading the corrected images to the HDFS distributed file system until all images to be corrected are corrected. This invention obtains the distortion calibration matrix of the target camera from a chessboard image and creates an RDD (Resilient Distributed Dataset) task based on the images to be corrected in the HDFS distributed file system. Using the distortion calibration matrix, a foreach operator is used to correct each image in the RDD task, resulting in a corrected image. This combines the distortion calibration matrix obtained from the chessboard image with large-scale image processing, achieving the goal of batch correction of a large number of distorted images simultaneously. This significantly improves the efficiency of image correction, reduces the cost, and facilitates the reuse of image data, lowering maintenance costs for administrators. Attached Figure Description

[0033] Figure 1 A flowchart of a distributed image distortion correction method provided by the present invention;

[0034] Figure 2A flowchart of an image correction method based on a Spark environment is provided for this invention;

[0035] Figure 3 This invention provides a schematic diagram of a distributed image distortion correction system.

[0036] Figure 4 A schematic diagram of the hardware structure of a possible electronic device provided by the present invention;

[0037] Figure 5 This is a schematic diagram of the hardware structure of a possible computer-readable storage medium provided by the present invention. Detailed Implementation

[0038] The specific embodiments of the present invention will be described in further detail below with reference to the accompanying drawings and examples. The following examples are for illustrative purposes only and are not intended to limit the scope of the invention.

[0039] Figure 1 A flowchart of a distributed image distortion correction method provided by the present invention is shown below. Figure 1 As shown, the method includes:

[0040] Step S100: Obtain the distortion calibration matrix of the target camera based on the chessboard image;

[0041] It should be noted that the executing entity of the method in this embodiment can be a computer terminal device with data processing, network communication, and program execution functions, such as a computer or a vehicle-mounted computer; it can also be a server device with the same or similar functions, or a cloud server with similar functions. This embodiment does not impose any restrictions on this. For ease of understanding, this embodiment and the following embodiments will be described using a server device as an example.

[0042] It is understood that the chessboard image mentioned above is a picture of a pre-prepared chessboard cardboard or other chessboard tool (a black and white chessboard) taken by the target camera mentioned above. This embodiment does not limit this. The target camera mentioned above is a fisheye camera that uses a fisheye lens.

[0043] It should be understood that, in order to further improve the accuracy of the above distortion calibration matrix, the above chessboard image can be an image taken from four lenses: front, back, left, and right of the above chessboard cardboard.

[0044] In the specific implementation, the target camera first takes pictures of the chessboard cardboard to obtain images from the front, back, left, and right sides. Then, the distortion calibration matrix of the target camera is obtained from the images through the visual processing algorithm.

[0045] Step S200: Create an RDD Resilient Distributed Dataset task based on multiple images to be corrected in the HDFS distributed file system;

[0046] It should be noted that the multiple images to be corrected in the HDFS distributed file system mentioned above can be images taken by the target camera mentioned above, or images taken by other cameras with the same parameters as the target camera mentioned above. This embodiment does not impose any restrictions on this.

[0047] Step S300: Use the foreach operator to correct each image to be corrected in the RDD Resilient Distributed Dataset task according to the distortion calibration matrix;

[0048] It should be noted that the foreach operator mentioned above can be a function in Spark used to operate on the aforementioned RDD Resilient Distributed Dataset task.

[0049] It is understandable that the above steps for image correction can be a process of recalculating each pixel block in the image to be corrected based on the above distortion calibration matrix and regenerating the corrected image.

[0050] Step S400: Upload the corrected images to the HDFS distributed file system until all images to be corrected are corrected.

[0051] Understandably, given the deficiencies in the background technology, this invention proposes a distributed image distortion correction method. The method includes: obtaining the distortion calibration matrix of the target camera based on a chessboard image; creating an RDD (Resilient Distributed Dataset) task based on multiple images to be corrected in the HDFS distributed file system; using a foreach operator to correct each image in the RDD task based on the distortion calibration matrix; and uploading the corrected images to the HDFS distributed file system until all images to be corrected are corrected. This invention obtains the distortion calibration matrix of the target camera from a chessboard image and creates an RDD (Resilient Distributed Dataset) task based on the images to be corrected in the HDFS distributed file system. Using the distortion calibration matrix, a foreach operator is used to correct each image in the RDD task, resulting in a corrected image. This combines the distortion calibration matrix obtained from the chessboard image with large-scale image processing, achieving the goal of batch correction of a large number of distorted images simultaneously. This significantly improves the efficiency of image correction, reduces the cost, and facilitates the reuse of image data, lowering maintenance costs for administrators.

[0052] In one possible embodiment, the step of obtaining the distortion calibration matrix of the target camera based on the chessboard image includes:

[0053] Step S101: Obtain four chessboard images from different angles: front, back, left, and right;

[0054] Step S102: Obtain the matrix mapping coefficients of the target camera using a preset visual processing algorithm and the chessboard image;

[0055] Step S103: Calculate the distortion calibration matrix of the target camera based on the matrix mapping coefficients.

[0056] In this embodiment, the matrix mapping coefficients of the target camera are obtained by taking four chessboard images from different angles (front, back, left, and right), thereby obtaining the distortion calibration matrix of the target camera. This greatly improves the accuracy of the distortion calibration matrix, and consequently, significantly improves the accuracy of the base table correction.

[0057] In one possible embodiment, the preset visual processing algorithm is an OpenCV-based visual processing algorithm.

[0058] In one possible embodiment, the step of correcting each image to be corrected in the RDD Resilient Distributed Dataset task using the foreach operator according to the distortion calibration matrix includes:

[0059] Step S301: Initialize the parameters of the preset visual processing algorithm according to the distortion calibration matrix;

[0060] Step S302: Use the Spark computing engine to correct each image to be corrected in the RDD Resilient Distributed Dataset task according to the initialized visual processing algorithm.

[0061] In this embodiment, the parameters of the visual processing algorithm are initialized using a distortion calibration matrix, and the images to be corrected in the RDD elastic distributed dataset task are corrected using the Spark big data computing engine based on the initialized visual processing algorithm. This fully utilizes the cluster advantage of the Spark big data computing engine for distributed processing, improving distortion correction efficiency. Compared with traditional applications or single-server programs, this method greatly improves efficiency, reduces time costs, and facilitates subsequent data reuse.

[0062] In one possible embodiment, after the step of calculating the distortion calibration matrix of the target camera based on the matrix mapping coefficients, the method includes:

[0063] Step S110: Obtain a preset free scaling factor, and optimize the distortion calibration matrix according to the preset free scaling factor.

[0064] It should be noted that the aforementioned free scaling factor can be the image Scala scaling factor, generally set to 0 or 1. When set to 0, the resulting distortion calibration matrix will crop out unwanted pixels as much as possible when applied to correct distorted images; when set to 1, the resulting distortion calibration matrix will introduce an extra black pixel and provide a ROI to crop it out.

[0065] In this embodiment of the invention, a free scaling factor is introduced to further optimize the distortion calibration matrix, thereby further improving the accuracy of the corrected image.

[0066] In one possible embodiment, the step of obtaining a preset free scaling factor and optimizing the distortion calibration matrix based on the preset free scaling factor includes:

[0067] Step S1101: When the preset free scaling factor is 0, the distortion calibration matrix is ​​pruned and optimized.

[0068] In one possible embodiment, the step of obtaining a preset free scaling factor and optimizing the distortion calibration matrix based on the preset free scaling factor further includes:

[0069] Step S1102: When the preset free scaling factor is 1, the distortion calibration matrix is ​​optimized by adding parameters.

[0070] In one possible application scenario, see Figure 2 , Figure 2 This invention provides a flowchart of an image correction method based on a Spark environment.

[0071] Spark refers to Apache Spark, a fast and general-purpose computing engine designed specifically for large-scale data processing. Resilient Distributed Datasets (RDDs) are an abstract concept in distributed computing. For developers, an RDD can be viewed as a Spark object that runs in memory.

[0072] In the above application scenario, the first step is to set up the basic Spark runtime environment. Then, connect to Spark to import data into the HDFS distributed file system. Based on the data files in the HDFS distributed file system, create an RDD task. Simultaneously, take four images of the chessboard (front, back, left, and right) using a target fisheye camera. Use the `findChessboardCorners` method of the OpenCV vision framework to find the chessboard corners, and calculate the internal distortion coefficients K and D of the target fisheye camera using `fisheye.calibrate`. Generate a distortion calibration matrix based on the internal distortion coefficients K and D, and optimize the distortion calibration matrix using Scala scaling factors for the images. Then, initialize parameters such as `map1` and `map2` based on the optimized distortion calibration matrix. Call Spark's `foreach` operator to calculate the RDD task based on the initialized parameters `map1` and `map2`, output the corrected image, and upload the corrected image data until all images are corrected.

[0073] Figure 3 This is a schematic diagram of a distributed image distortion correction system provided in an embodiment of the present invention, as shown below. Figure 3 As shown, a distributed image distortion correction system includes a distortion calculation module 100, a task creation module 200, an image correction module 300, and an image upload module 400, wherein:

[0074] The distortion calculation module 100 is used to obtain the intrinsic parameter matrix of the target camera based on the chessboard image; the task creation module 200 is used to create an RDD elastic distributed dataset task based on multiple images to be corrected in the HDFS distributed file system; the image correction module 300 is used to correct each image to be corrected in the RDD elastic distributed dataset task using the foreach operator based on the distortion calibration matrix; and the image upload module 400 is used to upload the corrected images to the HDFS distributed file system until all images to be corrected are corrected.

[0075] This invention provides a distributed image distortion correction system, which obtains the distortion calibration matrix of the target camera based on a chessboard image; creates an RDD (Resilient Distributed Dataset) task based on multiple images to be corrected in the HDFS distributed file system; uses a foreach operator to correct each image in the RDD task based on the distortion calibration matrix; and uploads the corrected images to the HDFS distributed file system until all images to be corrected are corrected. This invention obtains the distortion calibration matrix of the target camera from a chessboard image and creates an RDD (Resilient Distributed Dataset) task based on the images to be corrected in the HDFS distributed file system. Using the distortion calibration matrix, a foreach operator is used to correct each image in the RDD task, resulting in a corrected image. This combines the distortion calibration matrix obtained from the chessboard image with large-scale image processing, achieving the goal of batch correction of a large number of distorted images simultaneously. This significantly improves the efficiency of image correction, reduces the cost, and facilitates the reuse of image data, lowering maintenance costs for administrators.

[0076] It is understood that the distributed image distortion correction system provided by this invention corresponds to the distributed image distortion correction method provided in the foregoing embodiments. The relevant technical features of the distributed image distortion correction system can be referred to the relevant technical features of the distributed image distortion correction method, and will not be repeated here.

[0077] Please see Figure 4 , Figure 4 This is a schematic diagram illustrating an embodiment of the electronic device provided in this invention. For example... Figure 4 As shown, this embodiment of the invention provides an electronic device, including a memory 1310, a processor 1320, and a computer program 1311 stored in the memory 1310 and executable on the processor 1320. When the processor 1320 executes the computer program 1311, it performs the following steps:

[0078] Obtain the distortion calibration matrix of the target camera from the chessboard image; create an RDD elastic distributed dataset task from multiple images to be rectified in the HDFS distributed file system; use the foreach operator to correct each image to be rectified in the RDD elastic distributed dataset task according to the distortion calibration matrix; upload the corrected images to the HDFS distributed file system until all images to be rectified are corrected.

[0079] Please see Figure 5 , Figure 5 This is a schematic diagram illustrating an embodiment of a computer-readable storage medium provided by the present invention. (See diagram below.) Figure 5As shown, this embodiment provides a computer-readable storage medium 1400, on which a computer program 1411 is stored. When the computer program 1411 is executed by a processor, it performs the following steps:

[0080] Obtain the distortion calibration matrix of the target camera from the chessboard image; create an RDD elastic distributed dataset task from multiple images to be rectified in the HDFS distributed file system; use the foreach operator to correct each image to be rectified in the RDD elastic distributed dataset task according to the distortion calibration matrix; upload the corrected images to the HDFS distributed file system until all images to be rectified are corrected.

[0081] This invention provides a distributed image distortion correction method, system, electronic device, and storage medium. The method involves obtaining the distortion calibration matrix of the target camera based on a chessboard image; creating an RDD (Resilient Distributed Dataset) task from multiple images to be corrected in the HDFS distributed file system; using a foreach operator to correct each image in the RDD task based on the distortion calibration matrix; and uploading the corrected images to the HDFS distributed file system until all images to be corrected are corrected. This invention obtains the distortion calibration matrix of the target camera from a chessboard image and creates an RDD (Resilient Distributed Dataset) task based on the images to be corrected in the HDFS distributed file system. Using the distortion calibration matrix, a foreach operator is used to correct each image in the RDD task, resulting in a corrected image. This combines the distortion calibration matrix obtained from the chessboard image with large-scale image processing, achieving the goal of batch correction of a large number of distorted images simultaneously. This significantly improves the efficiency of image correction, reduces the cost, and facilitates the reuse of image data, lowering maintenance costs for administrators.

[0082] It should be noted that the descriptions of each embodiment in the above embodiments have different focuses. For parts that are not described in detail in a certain embodiment, please refer to the relevant descriptions in other embodiments.

[0083] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0084] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded computer, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0085] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0086] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0087] Although preferred embodiments of the invention have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including both the preferred embodiments and all changes and modifications falling within the scope of the invention.

[0088] Obviously, those skilled in the art can make various modifications and variations to this invention without departing from its spirit and scope. Therefore, if these modifications and variations fall within the scope of the claims of this invention and their equivalents, this invention also intends to include these modifications and variations.

Claims

1. A distributed picture distortion correction method, characterized by, The method comprises: obtaining a distortion calibration matrix of a target camera according to a chessboard picture; creating an RDD (Resilient Distributed Dataset) task according to a plurality of pictures to be corrected in an HDFS (Hadoop Distributed File System); correcting each picture to be corrected in the RDD task according to the distortion calibration matrix by using a foreach operator; uploading the corrected pictures to the HDFS until correction of all the pictures to be corrected is completed; and the step of obtaining the distortion calibration matrix of the target camera according to the chessboard picture comprises: obtaining four chessboard pictures at different angles, front, back, left and right; obtaining a matrix mapping coefficient of the target camera by using a preset visual processing algorithm and the chessboard picture; calculating the distortion calibration matrix of the target camera according to the matrix mapping coefficient; the step of correcting each picture to be corrected in the RDD task according to the distortion calibration matrix by using the foreach operator comprises: initializing parameters of the preset visual processing algorithm according to the distortion calibration matrix; and correcting each picture to be corrected in the RDD task according to the initialized visual processing algorithm by using a Spark computing engine.

2. The distributed picture distortion correction method of claim 1, wherein, The preset visual processing algorithm is an openCV-based visual processing algorithm.

3. The distributed picture distortion correction method of claim 1, wherein, after the step of calculating the distortion calibration matrix of the target camera according to the matrix mapping coefficient, the method further comprises: obtaining a preset free scale factor, and optimizing the distortion calibration matrix according to the preset free scale factor.

4. The distributed picture distortion correction method of claim 3, wherein, the step of obtaining the preset free scale factor and optimizing the distortion calibration matrix according to the preset free scale factor comprises: when the preset free scale factor is 0, performing cropping optimization on the distortion calibration matrix.

5. The distributed picture distortion correction method of claim 3, wherein, the step of obtaining the preset free scale factor and optimizing the distortion calibration matrix according to the preset free scale factor further comprises: when the preset free scale factor is 1, performing addition optimization on the distortion calibration matrix.

6. A distributed picture distortion correction system, characterized by, The method comprises a distortion calculation module configured to obtain a distortion calibration matrix of a target camera according to a chessboard picture; a task creation module configured to create an RDD (Resilient Distributed Dataset) task according to a plurality of pictures to be corrected in an HDFS (Hadoop Distributed File System); a picture correction module configured to correct each picture to be corrected in the RDD task according to the distortion calibration matrix by using a foreach operator; a picture uploading module configured to upload the corrected pictures to the HDFS until correction of all the pictures to be corrected is completed; and the distortion calibration matrix of the target camera obtained according to the chessboard picture comprises: obtaining four chessboard pictures at different angles, front, back, left and right; obtaining a matrix mapping coefficient of the target camera by using a preset visual processing algorithm and the chessboard picture; calculating the distortion calibration matrix of the target camera according to the matrix mapping coefficient; the distortion calculation module is configured to obtain a distortion calibration matrix of a target camera according to a chessboard picture; The rectification of each picture to be rectified in the RDD elastic distributed dataset task according to the distortion calibration matrix comprises: Parameter initialization of the preset visual processing algorithm according to the distortion calibration matrix; Rectification of each picture to be rectified in the RDD elastic distributed dataset task according to the initialized visual processing algorithm by using the Spark computing engine.

7. An electronic device, comprising: The distributed picture distortion rectification method comprises a memory and a processor, wherein the processor is configured to execute a computer management program stored in the memory to implement the steps of the distributed picture distortion rectification method according to any one of claims 1-5.

8. A computer-readable storage medium, characterized in that, The distributed picture distortion rectification method comprises a memory and a processor, wherein the processor is configured to execute a computer management program stored in the memory to implement the steps of the distributed picture distortion rectification method according to any one of claims 1-5.

Images