Method, device, medium, product and system for realizing panoramic image stitching

Abstract

The invention provides a method, a device, a medium, a product and a system for realizing panoramic image stitching. The method comprises the following steps: respectively receiving images of connected cameras through a heterogeneous platform; determining a global target brightness value and adjusting the aperture, the light sensitivity and the shutter speed of each camera according to the global target brightness value; determining a global white balance parameter value and adjusting an RGB channel corresponding to the image shot by each camera according to the global white balance parameter value; determining a global noise reduction parameter value and carrying out noise reduction processing on the image shot by each camera according to the global noise reduction parameter value; and splicing the processed images into a panoramic image. By means of the technical scheme, the problem existing in panoramic image splicing in the prior art is solved through fusion of heterogeneous ISP platforms, and a low-cost solution is provided especially for the scenes that an existing ISP platform cannot meet the requirement for the number of camera paths and the research and development cost of replacing the whole ISP platform is too high.

Description

Technical Field

[0001] The present invention relates to the technical field of panoramic image stitching, and in particular, to a method, device, medium, product, and system for implementing panoramic image stitching. Background Art

[0002] With the rapid development of virtual reality and augmented reality technologies, 360-degree panoramic images have been widely used in multiple fields, such as virtual tourism, real estate display, security monitoring, vehicle-mounted, etc. To implement a 360-degree panoramic image, multiple cameras are usually used, and then the images of each camera are stitched. Currently, generally, an image processing ISP platform with a corresponding number of cameras is selected according to the required number of cameras, and relevant automatic exposure control, white balance correction, noise reduction processing, and stitching algorithms are designed based on this ISP platform for image stitching. However, in some scenarios, there will be problems that the existing ISP platform cannot meet the requirements of the number of cameras, and the R & D cost of replacing the entire ISP platform is too high. Summary of the Invention

[0003] In view of the above problems in the prior art, embodiments of the present invention provide a method and device for panoramic image stitching implemented using a heterogeneous ISP platform.

[0004] To achieve the above object, on the one hand, a method for implementing panoramic image stitching is provided, which uses at least a first ISP platform and a second ISP platform of different types to implement panoramic image stitching. The method includes:

[0005] Receiving M images from M cameras through the first ISP platform and N images from N cameras through the second ISP platform. The M images respectively correspond to M predetermined orientations, and the N images respectively correspond to N predetermined orientations. Any two orientations among the M orientations and the N orientations are different. Here, M is an integer greater than or equal to 1, N is an integer greater than or equal to 1, and the M orientations plus the N orientations include at least four of the following orientations: front, rear, left, right, left front, left rear, right front, and right rear;

[0006] Brightness adjustment step: Obtaining a first platform brightness value of the M images through the first ISP platform and a second platform brightness value of the N images through the second ISP platform, determining a global target brightness value, and adjusting the aperture, sensitivity, and shutter speed of each camera according to the global target brightness value. Here, the global target brightness value is a weighted average of the first platform brightness value and the second platform brightness value according to a first predetermined weight ratio;

[0007] Color adjustment step: Obtain the first platform white balance parameter values of the M images through the first ISP platform and obtain the second platform white balance parameter values of the N images through the second ISP platform, determine the global white balance parameter value, and adjust the RGB channels corresponding to the images captured by each camera according to the global white balance parameter value, where the global white balance parameter value is the weighted average of the first platform white balance parameter value and the second platform white balance parameter value according to a second predetermined weight ratio;

[0008] Noise reduction step: Obtain the first platform noise reduction parameter values of the M images through the first ISP platform and obtain the second platform noise reduction parameter values of the N images through the second ISP platform, determine the global noise reduction parameter value, and perform noise reduction processing on the images captured by each camera according to the global noise reduction parameter value, where the global noise reduction parameter value is the weighted average of the first platform noise reduction parameter value and the second platform noise reduction parameter value according to a third predetermined weight ratio;

[0009] Stitching step: Detect feature points of the M images through the first ISP platform and detect feature points of the N images through the second ISP platform, use a predetermined feature point matching algorithm to determine the relative positions and poses between the images in the M images and the N images, and stitch the M images and the N images that have undergone brightness adjustment, color adjustment, and noise reduction into a panoramic image according to a predetermined image fusion algorithm.

[0010] Preferably, in the method, where:

[0011] The global target brightness value is the average of the first platform brightness value and the second platform brightness value;

[0012] The global white balance parameter value is the average of the first platform white balance parameter value and the second platform white balance parameter value;

[0013] The global noise reduction parameter value is the average of the first platform noise reduction parameter value and the second platform noise reduction parameter value.

[0014] Preferably, in the method, where the brightness adjustment step further includes:

[0015] Dynamically adjust the exposure parameter according to the brightness difference between each pair of adjacent images in the M images and the N images.

[0016] Preferably, in the method, where the color adjustment step further includes:

[0017] Dynamically adjust the white balance parameter according to the color difference between each pair of adjacent images in the M images and the N images.

[0018] Preferably, in the method, the noise reduction step further includes:

[0019] dynamically adjusting the noise reduction parameter according to the noise difference between each pair of adjacent images among the M images and the N images.

[0020] Preferably, in the method, the M orientations include four orientations: front, rear, left rear, and right rear; the N orientations include two orientations: left front and right front.

[0021] On the other hand, there is provided an apparatus for implementing panoramic image stitching, including a memory and a processor. The memory stores at least one program, and the at least one program is executed by the processor to implement the steps of the method for implementing panoramic image stitching as described above.

[0022] On another aspect, there is provided a computer-readable storage medium storing at least one program, and the at least one program is executed by the processor to implement the steps of the method for implementing panoramic image stitching as described above.

[0023] On yet another aspect, there is provided a computer program product including a computer program, characterized in that the computer program, when executed by a processor, implements the steps of the method for implementing panoramic image stitching as described above.

[0024] On yet another aspect, there is provided a system for implementing panoramic image stitching, which uses the method as described above to implement panoramic image stitching, including:

[0025] a first ISP platform connected to the M cameras respectively arranged in the M orientations;

[0026] a second ISP platform connected to the N cameras respectively arranged in the N orientations;

[0027] wherein the first ISP platform and the second ISP platform are heterogeneous and exchange information through a data bus; and,

[0028] the computer-readable storage medium as described above.

[0029] The above technical solution has the following technical effects:

[0030] By using multiple heterogeneous ISP platforms that work independently, the problems of inconsistent brightness, inconsistent color, and many noise points existing in panoramic image stitching in the prior art are solved by fusing the image processing results of multiple heterogeneous ISP platforms. This method can achieve automatic exposure control, white balance correction, noise reduction processing, and precise stitching of multiple images through the cooperation of multiple, such as two different ISP platforms; thus, the problem that the existing ISP platform cannot meet the demand for the number of camera channels in some scenarios and the R & D cost of replacing the entire ISP platform is too high is solved.

[0031] In a specific application, it is especially suitable for the transformation of existing systems. When the existing ISP platform cannot meet the requirements, the existing ISP platform can be used as the main platform, and another low-cost ISP platform can be added as an auxiliary ISP platform, and then the processing results of the two ISP platforms are fused, so that high-precision panoramic image stitching can be achieved at a lower cost. BRIEF DESCRIPTION OF THE DRAWINGS

[0032] Figure 1 FIG. is a schematic diagram of the cooperation structure of panoramic image stitching implemented using two heterogeneous ISP platforms according to an embodiment of the present invention;

[0033] Figure 2 FIG. is a schematic flowchart of a method for implementing panoramic image stitching according to an embodiment of the present invention;

[0034] Figure 3 FIG. is a schematic diagram of the principle of a method for implementing panoramic image stitching according to another embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0035] To further illustrate the embodiments, the present invention provides drawings. These drawings are a part of the disclosure of the present invention, which are mainly used to illustrate the embodiments and can be used to explain the operating principle of the embodiments in conjunction with the relevant descriptions in the specification. With reference to these contents, those of ordinary skill in the art should be able to understand other possible implementation manners and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are usually used to represent similar components.

[0036] The present invention will be further described below in conjunction with the drawings and specific embodiments.

[0037] Embodiment 1:

[0038] The present invention provides a method for implementing panoramic image stitching using at least a first ISP platform and a second ISP platform that are heterogeneous. Figure 1The collaborative structure diagram of panoramic image stitching using two heterogeneous ISP platforms according to an embodiment of the present invention is shown in FIG. The embodiment of the present invention uses two different ISP platforms to obtain multiple overlapping images respectively, and performs brightness adjustment, color adjustment, noise reduction processing and panoramic image stitching.

[0039] Figure 2 FIG. 1 is a flow chart of a method for realizing panoramic image stitching according to an embodiment of the present invention. Figure 2 The method for realizing panoramic image stitching in this embodiment includes the following steps:

[0040] Receiving M images from M cameras through a first ISP platform and receiving N images from N cameras through a second ISP platform, the M images correspond to predetermined M directions respectively, the N images correspond to predetermined N directions respectively, any two directions of the M directions are different from any two directions of the N directions, wherein M is an integer greater than or equal to 1, N is an integer greater than or equal to 1, and the M directions plus the N directions include at least four of the following directions: front, back, left, right, left front, left back, right front, and right back;

[0041] Brightness adjustment step: obtain the first platform brightness values ​​of M images through the first ISP platform and obtain the second platform brightness values ​​of N images through the second ISP platform, determine the global target brightness value and adjust the aperture, sensitivity and shutter speed of each camera according to the global target brightness value, wherein the global target brightness value is the weighted average of the first platform brightness value and the second platform brightness value according to the first predetermined weight ratio; in a specific implementation, the global target brightness value is the average of the first platform brightness value and the second platform brightness value; in other implementations, the first predetermined weight ratio is set according to the size of M and N, for example, when M is greater than N, the weight of the first platform brightness value is set to be greater than the weight of the second platform brightness value; in a specific implementation, the algorithm used by the first ISP platform and the second ISP platform to obtain the brightness value of the corresponding platform is the brightness value calculation method of the prior art, such as the brightness value calculation method of the ISP platform, which is not repeated here; automatic exposure control is achieved through this step;

[0042] Color adjustment step: obtaining white balance parameter values ​​of the first platform of M images through the first ISP platform and obtaining white balance parameter values ​​of the second platform of N images through the second ISP platform, determining a global white balance parameter value and adjusting the RGB channels corresponding to the images taken by each camera according to the global white balance parameter value, wherein the global white balance parameter value is a weighted average of the white balance parameter value of the first platform and the white balance parameter value of the second platform according to a second predetermined weight ratio; in a specific implementation, the global white balance parameter value is an average of the white balance parameter value of the first platform and the white balance parameter value of the second platform; in other implementations, the second predetermined weight ratio is set according to the size of M and N, for example, when M is greater than N, the weight of the white balance parameter value of the first platform is set to be greater than the weight of the white balance parameter value of the second platform; in a specific implementation, the algorithm used by the first ISP platform and the second ISP platform to obtain the white balance parameter values ​​of the corresponding platforms is a white balance value calculation method of the prior art, such as a white balance parameter value calculation method provided by the ISP platform, which is not described in detail here; wherein, color correction is performed on each image through the white balance parameter to ensure color consistency;

[0043] Noise reduction step: obtaining a first platform noise reduction parameter value of M images through a first ISP platform and obtaining a second platform noise reduction parameter value of N images through a second ISP platform, determining a global noise reduction parameter value and performing noise reduction processing on the image taken by each camera according to the global noise reduction parameter value, wherein the global noise reduction parameter value is a weighted average of the first platform noise reduction parameter value and the second platform noise reduction parameter value according to a third predetermined weight ratio; in a specific implementation, the global noise reduction parameter value is an average of the first platform noise reduction parameter value and the second platform noise reduction parameter value; in other implementations, the first predetermined weight ratio is set according to the size of M and N, for example, when M is greater than N, the weight of the first platform noise reduction parameter value is set to be greater than the weight of the second platform noise reduction parameter value; in a specific implementation, the algorithm used by the first ISP platform and the second ISP platform to obtain the noise reduction parameter values ​​of the corresponding platforms is a noise reduction parameter value calculation method of the prior art, such as the noise reduction parameter value calculation method provided by the ISP platform, which is not described in detail here; wherein, the image noise is reduced by noise reduction technology to improve the image quality;

[0044] Stitching steps: Detect feature points of M images through the first ISP platform and detect feature points of N images through the second ISP platform. Use a predetermined feature point matching algorithm to determine the relative positions and poses between each image in the M images and the N images, and stitch the M images and the N images that have undergone brightness adjustment, color adjustment, and noise reduction into a panoramic image according to a predetermined image fusion algorithm. In specific implementation, the first ISP platform and the second ISP platform use existing feature point detection methods to detect feature points of images, use existing feature point matching algorithms to determine the relative positions and poses between each image, and perform panoramic stitching of each image according to existing image fusion algorithms.

[0045] Embodiment 2:

[0046] In this embodiment, the first ISP platform is platform A and the second ISP platform is platform B; among them, platform A is connected to four cameras, namely the front, rear, left rear, and right rear cameras; platform B is connected to two cameras, namely the left front and right front cameras; as Figure 3 , the method of this embodiment performs brightness adjustment of images through an automatic exposure algorithm, performs color adjustment of images through a color algorithm module, performs noise reduction processing through a noise reduction algorithm module, and finally performs image stitching on the images that have undergone the foregoing processing.

[0047] Automatic exposure module: Platforms A and B respectively independently calculate the current brightness of their corresponding images using a predetermined automatic exposure algorithm, and exchange brightness information through a data bus; then, based on the brightness information of platforms A and B, calculate the global target brightness value, and the calculation method is, for example, the weighted average method mentioned in Embodiment 1. Next, each ISP adjusts its corresponding aperture, sensitivity, and shutter speed according to the global target brightness value. Specifically, the aperture controls the amount of light entering the lens, the sensitivity adjusts the sensitivity of the image sensor to light, and the shutter speed determines the exposure time. Through the coordinated adjustment of these three parameters, the automatic exposure algorithm can precisely control the brightness of each image to make it close to the target brightness value.

[0048] To ensure the brightness consistency between multiple images, the automatic exposure algorithm also introduces a brightness smooth transition mechanism. This mechanism analyzes the brightness differences between adjacent images and dynamically adjusts the exposure parameters, making the brightness change smoothly between images and avoiding obvious brightness jumps. In addition, the predetermined automatic exposure algorithm can also consider the dynamic range of the scene and ensure that details in both highlight and shadow areas are fully retained through local exposure adjustment.

[0049] Color algorithm module: Through four cameras on the A platform at the front, rear, left rear, and right rear, and two cameras on the B platform at the left front and right front, the ISP corresponding to each camera independently analyzes the color information of its corresponding image, identifies potential white or neutral gray areas, and obtains the white balance parameter values corresponding to the cameras on each platform. Then, color information is exchanged through the data bus, and the global white balance parameter value is calculated using the weighted average method mentioned in Embodiment 1. Next, each ISP adjusts the RGB channels of its corresponding image according to the global white balance parameter. Specifically, the color restoration algorithm adjusts the intensity of each color channel through gain control, so that the overall color balance of the image is close to the real scene.

[0050] To ensure color consistency between multiple images, the color restoration algorithm also introduces a color smooth transition mechanism. This mechanism dynamically adjusts the white balance parameter by analyzing the color differences between adjacent images, so that the color changes smoothly between images, avoiding obvious color differences. In addition, the color restoration algorithm also considers the lighting conditions of the scene, and through local color adjustment, ensures that the colors in different lighting areas can be accurately restored.

[0051] Noise reduction algorithm module: Through four cameras on the A platform at the front, rear, left rear, and right rear, and two cameras on the B platform at the left front and right front, the ISP corresponding to each camera independently analyzes the pixel values of its corresponding image, and uses a predetermined noise recognition algorithm to identify the position and intensity of the noise. Then, the noise information of each image is exchanged through the data bus, and a predetermined algorithm is used to calculate the global noise reduction parameter; for example, the weighted average method described above is used to calculate the global noise reduction parameter value. Next, each ISP performs noise reduction processing on its corresponding image according to the global noise reduction parameter. In a specific implementation, methods such as spatial filtering, time-domain filtering, or frequency-domain filtering are used to reduce the noise in the image.

[0052] To ensure noise reduction consistency between multiple images, the noise reduction algorithm also includes: a noise reduction smooth transition mechanism. This mechanism dynamically adjusts the noise reduction parameter by analyzing the noise differences between adjacent images, so that the noise reduction effect smoothly transitions between images, avoiding obvious noise reduction traces. In addition, the noise reduction algorithm also considers the detailed information of the scene, and through local noise reduction adjustment, ensures that the details and textures of the image can be fully retained. Through the above steps, the noise reduction algorithm realizes efficient noise reduction processing in two different ISP platforms, ensuring the clarity and high quality of the 360-degree panoramic image.

[0053] Image stitching: First, each ISP independently detects the feature points of its corresponding image and exchanges the feature point information through the data bus. Then, through the feature point matching algorithm, the relative positions and poses between the images are calculated. Next, according to the relative positions and poses, geometric transformation is performed on the images to align them in the same coordinate system. Finally, through the image fusion algorithm, the stitching traces are eliminated to generate a seamless panoramic image.

[0054] In the specific implementation of the embodiments of the present invention, the automatic exposure algorithm, color adjustment algorithm, noise recognition and noise reduction processing algorithm, and the final image stitching algorithm adopted by each ISP platform are all corresponding algorithms of the prior art. The improvement of the present invention mainly lies in the innovative use of multiple heterogeneous ISP platforms that work independently and fusing the processing results of multiple heterogeneous ISP platforms to solve the problem that in some scenarios, the existing ISP platforms cannot meet the requirements of the number of camera channels and the R & D cost of replacing the entire ISP platform is too high.

[0055] Embodiment Three:

[0056] The present invention also provides a device for implementing panoramic image stitching. The device includes a memory and a processor. The memory stores at least one segment of program, and the at least one segment of program is executed by the processor to implement the method for implementing panoramic image stitching as described in any one of the above.

[0057] Furthermore, as an executable solution, the so-called processor may be a Central Processing Unit (CPU), or may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field-Programmable Gate Arrays (FPGAs), or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. The general-purpose processor may be a microprocessor or the processor may also be any conventional processor, etc. The processor is the control center of the computer unit, connecting various parts of the entire computer unit through various interfaces and lines.

[0058] The memory can be used to store the computer program and / or modules. By running or executing the computer program and / or modules stored in the memory, and invoking the data stored in the memory, the processor realizes various functions of the computer unit. The memory mainly includes a program storage area and a data storage area. Among them, the program storage area can store an operating system and application programs required for at least one function; the data storage area can store data created according to the use of the mobile phone, etc. In addition, the memory may include high-speed random access memory, and may also include non-volatile memory, such as a hard disk, a memory, a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) card, a Flash Card, at least one magnetic disk storage device, a flash memory device, or other volatile solid-state storage devices.

[0059] Embodiment 4:

[0060] The present invention also provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, the steps of the above method in the embodiments of the present invention are realized.

[0061] If the modules / units integrated in the computer unit are implemented in the form of software function units and sold or used as independent products, they can be stored in a computer-readable storage medium. Based on such an understanding, to implement all or part of the processes in the above method embodiments of the present invention, it can also be completed by instructing relevant hardware through a computer program. The computer program can be stored in a computer-readable storage medium. When the computer program is executed by a processor, the steps of the above method embodiments can be realized. Among them, the computer program includes computer program code, and the computer program code can be in the form of source code, object code, executable file or some intermediate form, etc. The computer-readable medium can include: any entity or device capable of carrying the computer program code, a recording medium, a USB flash drive, a mobile hard disk, a magnetic disk, an optical disk, a computer memory, a read-only memory (ROM), a random access memory (RAM), and a software distribution medium, etc. It should be noted that the content included in the computer-readable medium can be appropriately increased or decreased according to the requirements of legislation and patent practice in the jurisdiction.

[0062] Embodiment 5:

[0063] The present invention also provides a computer program product including a computer program, and when the computer program is executed by a processor, the steps of the method as described above are realized.

[0064] Embodiment Six:

[0065] The present invention further provides a system for implementing panoramic image stitching, which uses the method described in any one of the above to implement panoramic image stitching. The system includes:

[0066] A first ISP platform, connected to M cameras respectively arranged in M directions;

[0067] A second ISP platform, connected to N cameras respectively arranged in N directions;

[0068] wherein, the first ISP platform and the second ISP platform are heterogeneous and exchange information through a data bus; and,

[0069] A computer-readable storage medium or a computer program product as described above.

[0070] Although the present invention is specifically shown and described in conjunction with the preferred embodiments, those skilled in the art should understand that various changes can be made to the present invention in form and detail without departing from the spirit and scope of the present invention defined by the appended claims, and all of them are within the protection scope of the present invention.

Claims

1. A method for realizing panoramic image stitching, characterized in that: Using at least a first ISP platform and a second ISP platform of heterogeneity to implement panoramic image stitching, the method includes: Receiving M images from M cameras through the first ISP platform and receiving N images from N cameras through the second ISP platform, the M images respectively correspond to predetermined M directions, the N images respectively correspond to predetermined N directions, the M directions are different from any two directions of the N directions, wherein M is an integer greater than or equal to 1, N is an integer greater than or equal to 1, and the M directions plus the N directions include at least four of the following directions: front, back, left, right, left front, left back, right front, and right back; Brightness adjustment step: obtaining first platform brightness values ​​of the M images through the first ISP platform and obtaining second platform brightness values ​​of the N images through the second ISP platform, determining a global target brightness value and adjusting the aperture, sensitivity and shutter speed of each camera according to the global target brightness value, wherein the global target brightness value is a weighted average of the brightness value of the first platform and the brightness value of the second platform according to a first predetermined weight ratio; Color adjustment step: obtaining a first platform white balance parameter value of the M images through the first ISP platform and obtaining a second platform white balance parameter value of the N images through the second ISP platform, determining a global white balance parameter value and adjusting the RGB channels corresponding to the image taken by each camera according to the global white balance parameter value, wherein the global white balance parameter value is a weighted average of the first platform white balance parameter value and the second platform white balance parameter value according to a second predetermined weight ratio; Noise reduction step: obtaining the first platform noise reduction parameter values ​​of the M images through the first ISP platform and obtaining the second platform noise reduction parameter values ​​of the N images through the second ISP platform, determining a global noise reduction parameter value and performing noise reduction processing on the image taken by each camera according to the global noise reduction parameter value, wherein the global noise reduction parameter value is a weighted average of the first platform noise reduction parameter value and the second platform noise reduction parameter value according to a third predetermined weight ratio; Stitching step: performing feature point detection on the M images through the first ISP platform and performing feature point detection on the N images through the second ISP platform, using a predetermined feature point matching algorithm to determine the relative position and posture between the M images and each of the N images, and stitching the M images and the N images that have undergone brightness adjustment, color adjustment and noise reduction into a panoramic image according to a predetermined image fusion algorithm.

2. The method according to claim 1, characterized in that: The global target brightness value is an average of the brightness value of the first platform and the brightness value of the second platform; The global white balance parameter value is an average of the first platform white balance parameter value and the second platform white balance parameter value; The global noise reduction parameter value is an average value of the first platform noise reduction parameter value and the second platform noise reduction parameter value.

3. The method according to claim 1, characterized in that The brightness adjustment step further includes: The exposure parameters are dynamically adjusted according to the brightness difference between each pair of adjacent images in the M images and the N images.

4. The method according to claim 1, characterized in that: The color adjustment step further comprises: The white balance parameter is dynamically adjusted according to the color difference between each pair of adjacent images in the M images and the N images.

5. The method according to claim 1, characterized in that The noise reduction step further comprises: The noise reduction parameters are dynamically adjusted according to the difference in noise points between each pair of adjacent images in the M images and the N images.

6. The method according to claim 1, characterized in that The M directions include four directions: front, back, left rear and right rear; and the N directions include two directions: left front and right front.

7. A device for realizing panoramic image stitching, characterized in that: The method comprises a memory and a processor, wherein the memory stores at least one program, and the at least one program is executed by the processor to implement the steps of the method for realizing panoramic image stitching as claimed in any one of claims 1 to 6.

8. A computer-readable storage medium, characterized in that: The storage medium stores at least one program, and the at least one program is executed by a processor to implement the steps of the method for realizing panoramic image stitching as described in any one of claims 1 to 6.

9. A computer program product, comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method for realizing panoramic image stitching as claimed in any one of claims 1 to 6 are implemented.

10. A system for realizing panoramic image stitching, characterized in that: Using the method described in any one of claims 1 to 6 to realize panoramic image stitching, comprising: A first ISP platform is connected to the M cameras respectively arranged at the M positions; A second ISP platform is connected to the N cameras respectively arranged at the N locations; The first ISP platform and the second ISP platform are heterogeneous and exchange information via a data bus; and The computer readable storage medium of claim 8.

Images