AI glasses image correction method and system based on factory calibration

Abstract

The invention belongs to the technical field of intelligent wearable equipment, and particularly relates to an AI glasses image correction method and system based on factory calibration, and the method comprises the steps: measuring and determining a fixed deviation angle of a camera optical axis direction relative to a standard human eye optical axis based on AI glasses, and taking the fixed deviation angle as an inherent calibration parameter of a calibration unit, storing the data into a nonvolatile memory of the AI glasses; starting and operating AI glasses, and acquiring an original image by using a camera of the AI glasses; reading inherent calibration parameters, and calculating geometric transformation parameters for correcting the original image in combination with internal parameters of the camera; geometric transformation processing is carried out on the original image by utilizing the geometric transformation parameters to generate a corrected image, the corrected image is taken as final shooting output, and by utilizing a method of combining factory accurate calibration and software compensation during operation, the'what you see is what you get 'shooting experience can be realized in a low-cost and low-power-consumption manner, the use satisfaction degree of a user is improved, and the user experience is improved. Meanwhile, implementation and maintenance are easy.

Description

Technical Field

[0001] This invention belongs to the field of smart wearable device technology, specifically relating to an AI glasses image correction method and system based on factory calibration. Background Technology

[0002] As an emerging smart terminal, one of the core functions of AI glasses is to capture the scene seen by the user. However, due to the limitations of the physical installation position of the camera, there is a fixed geometric deviation between its optical path and the natural visual axis of the human eye. For example, when the camera is usually installed on the side or upper middle part of the frame, there is a geometric deviation between its shooting angle and the angle of natural human observation. This deviation often makes it impossible for users to accurately predict the final image effect when taking pictures. The captured image differs from the angle or content actually seen by the human eye, which seriously affects the user experience.

[0003] Currently, although some image calibration methods exist, these methods are generally unable to effectively and cost-effectively solve the problem of fixed geometric deviations between the camera and the human eye. Summary of the Invention

[0004] This invention provides an AI glasses image correction method and system based on factory calibration. By calculating the inherent calibration parameters of the AI ​​glasses, the system can accurately calibrate the newly manufactured AI glasses and store them in the AI ​​glasses. During subsequent operation, geometric transformation parameters are calculated to facilitate real-time software compensation. By combining factory accurate calibration with runtime software compensation, a "what you see is what you get" photography experience can be achieved in a low-cost and low-power manner, improving user satisfaction. At the same time, it ensures the systematic nature and repeatability of the calibration process and is easy to implement and maintain.

[0005] An image correction method for AI glasses based on factory calibration includes: Based on the AI ​​glasses, the fixed deviation angle of the optical axis of its camera relative to the standard human eye optical axis is measured and determined, and the fixed deviation angle is used as the inherent calibration parameter of the calibration unit and stored in the non-volatile memory of the AI ​​glasses. Launch and run the AI ​​glasses, and use their camera to capture raw images; Read the inherent calibration parameters and combine them with the camera's internal parameters to calculate the geometric transformation parameters for correcting the original image; The original image is geometrically transformed using geometric transformation parameters to generate a corrected image, which is then used as the final output.

[0006] By calculating the inherent calibration parameters of the AI ​​glasses, the newly manufactured AI glasses can be accurately calibrated and stored in the AI ​​glasses. During subsequent operation, geometric transformation parameters are calculated to facilitate real-time software compensation. By combining accurate factory calibration and runtime software compensation, a "what you see is what you get" photography experience can be achieved in a low-cost and low-power manner, improving user satisfaction. At the same time, the systematic nature and repeatability of the calibration process are guaranteed, and it is easy to implement and maintain.

[0007] Furthermore, the step of measuring and determining a fixed deviation angle of the camera optical axis direction relative to the standard human eye optical axis based on the AI ​​glasses, and storing this fixed deviation angle as an inherent calibration parameter of the calibration unit in the memory of the AI ​​glasses, includes: Based on the AI ​​glasses to be calibrated, they are fixed on a three-dimensional fine-tuning fixture. A standard human eye optical axis is established using a laser alignment instrument. The posture of the AI ​​glasses to be calibrated is adjusted and made parallel to the calibration target plate. The AI ​​glasses' camera is activated to capture an image containing the calibration target, and the geometric center point of the calibration target in the image is detected to obtain its pixel coordinates. The pixel offset is obtained by comparing the pixel coordinates of the geometric center point with the pixel coordinates of the ideal geometric center point. Based on pixel offset and camera focal length, a fixed deviation angle of the camera's optical axis relative to the standard human eye's optical axis is calculated and determined; the fixed deviation angle includes a horizontal deviation angle and a vertical deviation angle. The fixed deviation angle is used as the inherent calibration parameter of the calibration unit, encoded in floating-point format, written into the non-volatile memory of the AI ​​glasses, and the current AI glasses are marked as having completed calibration.

[0008] High precision and repeatability of the calibration environment are ensured by using a 3D fine-tuning fixture and a laser alignment instrument; by calculating the pixel offset and determining the fixed deviation angle in combination with the camera focal length, the complex alignment problem can be transformed into an image processing problem, which is convenient for measurement, avoids subjective errors of manual operation, and improves parameter consistency during mass production; the floating-point number format is used for encoding and storage, which makes it easy to accurately retain data details while ensuring that data is not lost.

[0009] Furthermore, the step of fixing the AI ​​glasses to be calibrated onto a three-dimensional fine-tuning fixture, establishing a standard human eye optical axis using a laser alignment instrument, and adjusting the posture of the AI ​​glasses to be calibrated to be parallel with the calibration target plate includes: Fix the AI ​​glasses to be calibrated onto the 3D fine-tuning fixture, and start the calibration software to establish a connection between the AI ​​glasses to be calibrated and the laser alignment instrument; Turn on the laser alignment device, emit a collimated laser beam to define the direction of the standard human eye optical axis, and adjust the collimated laser beam to pass through the center point of the calibration target plate; The camera of the AI ​​glasses to be calibrated is activated to acquire the image of the calibration target board in real time. The corner points of the calibration target board image are extracted using a sub-pixel corner detection algorithm and the center point of the calibration target board image is calculated. The posture of the AI ​​glasses to be calibrated is adjusted using a fine-tuning fixture so that the center point of the calibration target board image is close to the ideal geometric center point. The direction of the collimated laser beam is adjusted and projected onto the outer surface of the lens of the AI ​​glasses to be calibrated to form a clear spot. The spot is positioned at the projection position of the center of the human eye pupil onto the lens of the AI ​​glasses to be calibrated, so as to achieve alignment between the standard human eye optical axis and the optical center of the system.

[0010] By employing a three-dimensional fine-tuning fixture, the posture adjustment error of the AI ​​glasses can be reduced. At the same time, the sub-pixel corner detection algorithm is used to reduce image detection error, thereby improving the alignment accuracy between the standard human eye optical axis and the system's optical center. Through dual calibration by adjusting the projection position of the collimating laser beam passing through the center of the calibration target plate and the light spot projected onto the center of the human eye's pupil, coarse alignment and precise calibration are achieved step by step, ensuring the accuracy of the measurement of fixed deviation angles. The calibration process is automated and requires no manual intervention, which can improve the efficiency of mass production and reduce the cost of manual calibration.

[0011] Furthermore, the expression for calculating the pixel offset is: ; ; In the formula, Indicates the horizontal pixel offset; The x-axis pixel coordinates representing the geometric center point; The x-axis pixel coordinates represent the ideal geometric center point; Indicates the vertical pixel offset; The vertical pixel coordinates of the geometric center point; The vertical pixel coordinates represent the ideal geometric center point; The formula for calculating the fixed deviation angle is: ; ; In the formula, Indicates the horizontal deviation angle; Indicates the vertical deviation angle; Represents the arctangent function; This indicates the camera's focal length.

[0012] Furthermore, it also includes: When the camera of AI glasses has radial distortion, the pixel coordinates of the geometric center point need to be corrected for distortion before calculating the pixel offset; The expression for distortion correction is: ; In the formula, The x-axis pixel coordinates represent the geometric center point after distortion correction; The vertical pixel coordinates of the geometric center point after distortion correction; The squared distance between the pixel coordinates of the geometric center point and the pixel coordinates of the ideal geometric center point is expressed as: ; Indicates the first distortion coefficient. This represents the second distortion coefficient.

[0013] By correcting the radial distortion of the AI ​​glasses camera, the impact of radial distortion on measurement accuracy can be eliminated.

[0014] Furthermore, the step of reading the inherent calibration parameters, combining them with the camera's internal parameters, and calculating the geometric transformation parameters for correcting the original image includes: Read the inherent parameters and obtain the camera's internal parameters; the camera's internal parameters include focal length and principal point coordinates; By combining inherent parameters and camera internal parameters, geometric transformation parameters are calculated to map the original image's viewpoint center from the camera's optical axis to the standard human eye's optical axis; these geometric transformation parameters include pixel value translation parameters or transformation matrices.

[0015] By calculating the geometric transformation parameters used to map the center of the original image's field of view from the camera's optical axis to the standard human eye's optical axis, it is possible to ensure that the center of the corrected image's field of view is completely consistent with the center of the human eye's field of view, thus solving the problem of discrepancies between shooting and human eye observation and improving shooting accuracy. At the same time, through various forms of geometric transformation parameters, it is possible to adapt to the performance requirements of different AI glasses, balancing correction accuracy and operating efficiency.

[0016] Furthermore, the calculation of the geometric transformation parameters also includes field-of-view matching correction, which corrects the geometric transformation parameters based on the difference between the standard human eye field of view and the camera field of view, so that the field of view of the corrected image matches human visual perception.

[0017] By using field-of-view matching correction, the captured images can be made to more closely resemble the actual scene observed by the human eye.

[0018] Furthermore, it also includes edge filling or content-aware filling of the corrected image to compensate for the missing image edges caused by geometric transformation processing.

[0019] By applying edge filling or content-aware filling to the corrected image, the integrity of the output image can be guaranteed, and the aesthetics of the captured image can be improved.

[0020] A system for image correction of AI glasses based on factory calibration includes: An image acquisition module includes a camera; the camera is integrated into the frame of the AI ​​glasses and is used to acquire raw images. The storage module includes a non-volatile memory for storing the inherent calibration parameters of the calibration unit; the inherent calibration parameters include a horizontal deviation angle and a vertical deviation angle. A calibration parameter reading module, which is connected to the storage module, is used to read the inherent calibration parameters of the AI ​​glasses; The calibration parameter calculation module, which is connected to the image acquisition module and the calibration parameter reading module, is used to calculate the geometric transformation parameters for correcting the original image based on the inherent calibration parameters and the camera's internal parameters. The image correction module, which connects the image acquisition module and the calibration parameter calculation module, is used to perform geometric transformation processing on the original image using geometric transformation parameters to generate a corrected image.

[0021] Furthermore, it also includes: The field-of-view matching unit, which is integrated into the calibration parameter calculation module, is used to correct the geometric transformation parameters based on the difference between the standard human eye field of view and the camera field of view. An image filling module, which is connected to the image correction module, is used to perform edge filling or content-aware filling on the corrected image to generate a complete image.

[0022] The beneficial effects of this invention are as follows: This invention calculates the inherent calibration parameters of AI glasses to accurately calibrate newly manufactured AI glasses and stores them in the AI ​​glasses. During subsequent operation, it calculates geometric transformation parameters to facilitate real-time software compensation. By combining accurate factory calibration with runtime software compensation, it can achieve a "what you see is what you get" photography experience in a low-cost and low-power manner, improving user satisfaction. At the same time, it ensures the systematicness and repeatability of the calibration process and is easy to implement and maintain. Attached Figure Description

[0023] Figure 1 This is a flowchart of the present invention; Figure 2 This is a schematic diagram of the system structure of the present invention; Figure 3 This is a schematic diagram of a computer device structure. Detailed Implementation

[0024] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0025] It should be noted that various aspects of embodiments within the scope of the appended claims are described below. It will be apparent that the aspects described herein can be embodied in a wide variety of forms, and any particular structure and / or function described herein is merely illustrative. Based on this disclosure, those skilled in the art will understand that one aspect described herein can be implemented independently of any other aspect, and two or more of these aspects can be combined in various ways. For example, any number of aspects set forth herein can be used to implement the device and / or practice the method. Additionally, this device and / or method can be implemented using other structures and / or functionalities besides one or more of the aspects set forth herein.

[0026] In addition, specific details are provided in the following description to facilitate a thorough understanding of the examples, and those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0027] Example 1 Figure 1 This illustrates an image correction method for AI glasses based on factory calibration. It calculates the inherent calibration parameters of the AI ​​glasses to accurately calibrate them upon leaving the factory, storing these parameters within the glasses. During subsequent runtime, it calculates geometric transformation parameters for real-time software compensation. This combination of factory calibration and runtime software compensation enables a "what you see is what you get" photography experience at low cost and low power consumption, improving user satisfaction. Simultaneously, it ensures the systematic nature and repeatability of the calibration process, making it easy to implement and maintain. Specifically, it includes the following steps: S1: Based on the AI ​​glasses, measure and determine the fixed deviation angle of its camera optical axis direction relative to the standard human eye optical axis, and store the fixed deviation angle as the inherent calibration parameter of the calibration unit in the non-volatile memory of the AI ​​glasses; S11: Based on the AI ​​glasses to be calibrated, fix them on the three-dimensional fine-tuning fixture, use a laser alignment instrument to establish the standard human eye optical axis, adjust the posture of the AI ​​glasses to be calibrated and make them parallel to the calibration target plate; S111: Fix the AI ​​glasses to be calibrated onto the 3D fine-tuning fixture, and start the calibration software on the host to establish a connection between the AI ​​glasses to be calibrated and the laser alignment instrument; In this embodiment, the three-dimensional fine-tuning fixture supports fine-tuning of XYZ three-axis translation and six degrees of freedom of pitch, yaw and roll, and the fine-tuning accuracy is no greater than 0.01°.

[0028] In this embodiment, the laser alignment device is a high-precision laser alignment device used to emit visible or near-infrared collimated laser beams. Its accuracy error is less than 0.01 mrad, and it is calibrated periodically.

[0029] S112: Activate the laser alignment device and emit a collimating laser beam to define the standard human eye optical axis direction. And adjust the collimated laser beam to pass through the center point of the calibration target plate to ensure that the optical axis of the standard human eye coincides with the normal of the calibration target plate; In this embodiment, the calibration target plate is provided with a high-precision checkerboard or dot array pattern, and the distance between it and the laser alignment instrument is greater than 2m.

[0030] S113: Activate the camera of the AI ​​glasses to be calibrated, acquire images of the calibration target plate in real time, extract corner points of the calibration target plate image using a sub-pixel corner detection algorithm, and calculate the center point of the calibration target plate image. The posture of the AI ​​glasses to be calibrated is adjusted using a fine-tuning fixture to center the image of the calibration target plate. Approaching the ideal geometric center point; In this embodiment, subpixel corner detection can be achieved using the cv2.cornerSubPix function in the OpenCV library.

[0031] In this embodiment, the pixel coordinates of the ideal geometric center point are: , , , Indicates the image width. Indicates the image height.

[0032] S114: Adjust the direction of the collimated laser beam and project it onto the outer surface of the lens of the AI ​​glasses to be calibrated to form a clear spot. Make the spot located at the projection position of the center of the human eye pupil onto the lens of the AI ​​glasses to be calibrated, so as to achieve alignment between the standard human eye optical axis and the optical center of the system. It should be noted that the projection position of the center of the human pupil on the lens of the AI ​​glasses to be calibrated refers to the position directly opposite the pupil when the human eye observes along the standard human eye optical axis. This is used to keep the collimating laser optical axis precisely aligned with the projection point of the center of the human pupil on the outer surface of the AI ​​glasses lens, so as to achieve alignment between the standard human eye optical axis and the optical center of the system. Thus, the direction of the collimating laser beam at this time is the direction of the standard human eye optical axis.

[0033] In practical applications, calibration is performed in a dark room to avoid ambient light interference and ensure that the collimated laser beam and the calibration target are clearly visible.

[0034] S12: Activate the AI ​​glasses' camera to capture an image containing the calibration target plate. And detect the geometric center point of the calibration target in the image. Get its pixel coordinates ; S13: Compare geometric center points pixel coordinates pixel coordinates of the ideal geometric center point , obtain pixel offset ; Among them, pixel offset The calculation expression is: ; ; In the formula, Indicates the horizontal pixel offset; The x-axis pixel coordinates representing the geometric center point; The x-axis pixel coordinates represent the ideal geometric center point; Indicates the vertical pixel offset; The vertical pixel coordinates of the geometric center point; The vertical pixel coordinates represent the ideal geometric center point; In this embodiment, when the AI ​​glasses' camera has radial distortion, the pixel coordinates of the geometric center point need to be determined before calculating the pixel offset. Distortion correction is performed using the pixel coordinates of the geometric center point after distortion correction. Calculate the pixel offset to eliminate the impact of radial distortion on measurement accuracy. The expression for distortion correction is as follows: ; In the formula, The x-axis pixel coordinates represent the geometric center point after distortion correction; The vertical pixel coordinates of the geometric center point after distortion correction; The squared distance between the pixel coordinates of the geometric center point and the pixel coordinates of the ideal geometric center point is expressed as: ; Indicates the first distortion coefficient. This represents the second distortion coefficient.

[0035] S14: Based on pixel offset Combined with camera focal length Calculate and determine the fixed deviation angle of the camera's optical axis direction relative to the standard human eye's optical axis. ; In this embodiment, the fixed deviation angle includes the horizontal deviation angle. and vertical deviation angle .

[0036] The formula for calculating the fixed deviation angle is as follows: ; ; In the formula, Indicates the horizontal deviation angle; Indicates the vertical deviation angle; Represents the arctangent function; Indicates the camera's focal length; In this embodiment, the horizontal deviation angle and vertical deviation angle The measurement accuracy reaches ±0.1°. When the camera's optical axis deflects to the right, then... , When the camera's optical axis is deflected upwards, then , .

[0037] S15: The fixed deviation angle is used as the inherent calibration parameter of the calibration unit, encoded in floating-point format, written into the non-volatile memory of the AI ​​glasses, and the current AI glasses are marked as having completed calibration. In this embodiment, the floating-point format includes IEEE 754; the non-volatile memory includes EEPROM and Flash, and the key name for writing is Calibration_Params.

[0038] S2: Start and run the AI ​​glasses, using their camera to capture raw images. ; S3: Read the inherent calibration parameters, combine them with the camera's internal parameters, and calculate the correction for the original image. Geometric transformation parameters; S31: Read inherent parameters and obtain internal parameters of the camera; In this embodiment, the camera's internal parameters include focal length. and principal point coordinates .

[0039] S32: Combining inherent parameters and camera internal parameters, calculate the geometric transformation parameters used to map the original image's viewpoint center from the camera's optical axis to the standard human eye's optical axis; In this embodiment, the geometric transformation parameters include pixel value translation parameters. Or the transformation matrix M.

[0040] In this embodiment, the calculation of geometric transformation parameters also includes field-of-view matching correction, which corrects the geometric transformation parameters based on the difference between the standard human eye field of view and the camera field of view, so that the field of view of the corrected image matches human visual perception, thereby making the captured image closer to the actual scene observed by the human eye.

[0041] S4: Perform geometric transformation on the original image using geometric transformation parameters to generate the corrected image. And use it as the final shooting output; In this embodiment, the corrected image is also included. Edge filling or content-aware filling is performed to compensate for image edge loss caused by geometric transformation processing, so as to ensure the integrity of the captured output image.

[0042] Example 2 Based on the same design concept, such as Figure 2 As shown, this embodiment provides an AI glasses image correction system based on factory calibration, including an image acquisition module, a storage module, a calibration parameter reading module, a calibration parameter calculation module, an image correction module, and an image filling module.

[0043] Specifically, the image acquisition module includes a camera; the camera is integrated into the frame of the AI ​​glasses and is used to acquire raw images; Specifically, the storage module includes non-volatile memory for storing the inherent calibration parameters of the calibration unit; the inherent calibration parameters include the horizontal deviation angle and the vertical deviation angle. Specifically, the calibration parameter reading module is connected to the storage module and is used to read the inherent calibration parameters of the AI ​​glasses; The calibration parameter calculation module integrates a field-of-view matching unit, which is used to correct the geometric transformation parameters based on the difference between the standard human eye field of view and the camera field of view.

[0044] Specifically, the calibration parameter calculation module, which connects the image acquisition module and the calibration parameter reading module, is used to calculate the geometric transformation parameters for correcting the original image based on the inherent calibration parameters and the camera's internal parameters. Specifically, the image correction module connects the image acquisition module and the calibration parameter calculation module. It is used to perform geometric transformation processing on the original image using geometric transformation parameters to generate a corrected image.

[0045] The image filling module, which is connected to the image correction module, is used to perform edge filling or content-aware filling on the corrected image to generate a complete image.

[0046] Example 3 Based on the same technical concept, embodiments of this application also provide a computer device, including a memory 1 and a processor 2, such as... Figure 3 As shown, memory 1 stores a computer program, and processor 2 executes the computer program to implement any of the methods described above.

[0047] The memory 1 includes at least one type of readable storage medium, including flash memory, hard disk, multimedia card, card-type memory (e.g., SD or DX memory), magnetic memory, magnetic disk, optical disk, etc. In some embodiments, the memory 1 can be an internal storage unit of the factory-calibrated AI glasses image correction system, such as a hard disk. In other embodiments, the memory 1 can also be an external storage device of the factory-calibrated AI glasses image correction system, such as a plug-in hard disk, smart media card (SMC), secure digital (SD) card, flash card, etc. Furthermore, the memory 1 can include both internal storage units and external storage devices of the factory-calibrated AI glasses image correction system. The memory 1 can be used not only to store application software and various data installed in the factory-calibrated AI glasses image correction system, such as the code of the factory-calibrated AI glasses image correction system program, but also to temporarily store data that has been output or will be output.

[0048] In some embodiments, processor 2 may be a central processing unit (CPU), controller, microcontroller, microprocessor or other data processing chip, used to run program code stored in memory 1 or process data, such as executing an AI glasses image correction system program based on factory calibration.

[0049] The present invention also discloses a computer-readable storage medium storing a computer program, which, when executed by a processor, performs the steps of the method described in the above-described method embodiments. The storage medium may be a volatile or non-volatile computer-readable storage medium.

[0050] The computer program product of the AI ​​glasses image correction method based on factory calibration provided in the embodiments of the present invention includes a computer-readable storage medium storing program code. The instructions included in the program code can be used to execute the steps of the method in the above method embodiments. For details, please refer to the above method embodiments, which will not be repeated here.

[0051] The present invention also discloses a computer program that, when executed by a processor, implements any of the methods described in the foregoing embodiments. This computer program product can be implemented specifically through hardware, software, or a combination thereof. In one optional embodiment, the computer program product is specifically embodied in a computer storage medium; in another optional embodiment, the computer program product is specifically embodied in a software product, such as a software development kit (SDK), etc.

[0052] It is understood that the same or similar parts in the above embodiments can be referred to each other, and the contents not described in detail in some embodiments can be referred to the same or similar contents in other embodiments.

[0053] It should be noted that in the description of this invention, the terms "first," "second," etc., are used for descriptive purposes only and should not be construed as indicating or implying relative importance. Furthermore, in the description of this invention, unless otherwise stated, "a plurality of" means at least two.

[0054] Any process or method description in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process, and the scope of the preferred embodiments of the invention includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functions involved, as will be understood by those skilled in the art to which embodiments of the invention pertain.

[0055] It should be understood that various parts of the present invention can be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. For example, if implemented in hardware, as in another embodiment, it can be implemented using any one or a combination of the following techniques known in the art: discrete logic circuits having logic gates for implementing logical functions on data signals, application-specific integrated circuits (ASICs) having suitable combinational logic gates, programmable gate arrays (PGAs), field-programmable gate arrays (FPGAs), etc.

[0056] Those skilled in the art will understand that all or part of the steps of the methods described in the above embodiments can be implemented by a program instructing related hardware, and the program can be stored in a computer-readable storage medium. When executed, the program includes one or a combination of the steps of the method embodiments.

[0057] Furthermore, the functional units in the various embodiments of the present invention can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium.

[0058] The storage media mentioned above can be read-only memory, disk, or optical disk, etc.

[0059] In the description of this specification, references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of the invention. In this specification, the illustrative expressions of the above terms do not necessarily refer to the same embodiment or example. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples.

[0060] Although embodiments of the present invention have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present invention. Those skilled in the art can make changes, modifications, substitutions and variations to the above embodiments within the scope of the present invention.

Claims

1. A method for image correction of AI glasses based on factory calibration, characterized in that, include: Based on the AI ​​glasses, the fixed deviation angle of the optical axis of its camera relative to the standard human eye optical axis is measured and determined, and the fixed deviation angle is used as the inherent calibration parameter of the calibration unit and stored in the non-volatile memory of the AI ​​glasses. Launch and run the AI ​​glasses, and use their camera to capture raw images; Read the inherent calibration parameters and combine them with the camera's internal parameters to calculate the geometric transformation parameters for correcting the original image; The original image is geometrically transformed using geometric transformation parameters to generate a corrected image, which is then used as the final output.

2. The AI ​​glasses image correction method based on factory calibration according to claim 1, characterized in that, The method based on AI glasses, measuring and determining a fixed deviation angle of its camera optical axis direction relative to the standard human eye optical axis, and storing this fixed deviation angle as an inherent calibration parameter of the calibration unit in the AI ​​glasses' memory, includes: Based on the AI ​​glasses to be calibrated, they are fixed on a three-dimensional fine-tuning fixture. A standard human eye optical axis is established using a laser alignment instrument. The posture of the AI ​​glasses to be calibrated is adjusted and made parallel to the calibration target plate. The AI ​​glasses' camera is activated to capture an image containing the calibration target, and the geometric center point of the calibration target in the image is detected to obtain its pixel coordinates. The pixel offset is obtained by comparing the pixel coordinates of the geometric center point with the pixel coordinates of the ideal geometric center point. Based on pixel offset and camera focal length, a fixed deviation angle of the camera's optical axis relative to the standard human eye's optical axis is calculated and determined; the fixed deviation angle includes a horizontal deviation angle and a vertical deviation angle. The fixed deviation angle is used as the inherent calibration parameter of the calibration unit, encoded in floating-point format, written into the non-volatile memory of the AI ​​glasses, and the current AI glasses are marked as having completed calibration.

3. The AI ​​glasses image correction method based on factory calibration according to claim 2, characterized in that, The process involves fixing the AI ​​glasses to be calibrated onto a three-dimensional fine-tuning fixture, establishing a standard human eye optical axis using a laser alignment instrument, and adjusting the posture of the AI ​​glasses to be calibrated to be parallel to the calibration target plate, including: Fix the AI ​​glasses to be calibrated onto the 3D fine-tuning fixture, and start the calibration software to establish a connection between the AI ​​glasses to be calibrated and the laser alignment instrument; Turn on the laser alignment device, emit a collimated laser beam to define the direction of the standard human eye optical axis, and adjust the collimated laser beam to pass through the center point of the calibration target plate; The camera of the AI ​​glasses to be calibrated is activated to acquire the image of the calibration target board in real time. The corner points of the calibration target board image are extracted using a sub-pixel corner detection algorithm and the center point of the calibration target board image is calculated. The posture of the AI ​​glasses to be calibrated is adjusted using a fine-tuning fixture so that the center point of the calibration target board image is close to the ideal geometric center point. The direction of the collimated laser beam is adjusted and projected onto the outer surface of the lens of the AI ​​glasses to be calibrated to form a clear spot. The spot is positioned at the projection position of the center of the human eye pupil onto the lens of the AI ​​glasses to be calibrated, so as to achieve alignment between the standard human eye optical axis and the optical center of the system.

4. The AI ​​glasses image correction method based on factory calibration according to claim 2, characterized in that, The expression for calculating the pixel offset is: ； ； In the formula, Indicates the horizontal pixel offset; The x-axis pixel coordinates representing the geometric center point; The x-axis pixel coordinates represent the ideal geometric center point; Indicates the vertical pixel offset; The vertical pixel coordinates of the geometric center point; The vertical pixel coordinates represent the ideal geometric center point; The formula for calculating the fixed deviation angle is: ； ； In the formula, Indicates the horizontal deviation angle; Indicates the vertical deviation angle; Represents the arctangent function; This indicates the camera's focal length.

5. The AI ​​glasses image correction method based on factory calibration according to claim 2, characterized in that, Also includes: When the camera of AI glasses has radial distortion, the pixel coordinates of the geometric center point need to be corrected for distortion before calculating the pixel offset; The expression for distortion correction is: ； In the formula, The x-axis pixel coordinates represent the geometric center point after distortion correction; The vertical pixel coordinates of the geometric center point after distortion correction; The expression for the squared distance between the pixel coordinates of the geometric center point and the pixel coordinates of the ideal geometric center point is: ; Indicates the first distortion coefficient. This represents the second distortion coefficient.

6. The AI ​​glasses image correction method based on factory calibration according to claim 1, characterized in that, The process of reading the inherent calibration parameters, combining them with the camera's internal parameters, and calculating the geometric transformation parameters for correcting the original image includes: Read the inherent parameters and obtain the camera's internal parameters; the camera's internal parameters include focal length and principal point coordinates; By combining inherent parameters and camera internal parameters, geometric transformation parameters are calculated to map the original image's viewpoint center from the camera's optical axis to the standard human eye's optical axis; these geometric transformation parameters include pixel value translation parameters or transformation matrices.

7. The AI ​​glasses image correction method based on factory calibration according to claim 1, characterized in that, The calculation of the geometric transformation parameters also includes field-of-view matching correction, which corrects the geometric transformation parameters based on the difference between the standard human eye field of view and the camera field of view, so that the field of view of the corrected image matches human visual perception.

8. The AI ​​glasses image correction method based on factory calibration according to claim 1, characterized in that, It also includes edge filling or content-aware filling of the corrected image to compensate for the loss of image edges caused by geometric transformation processing.

9. A system for implementing the AI ​​glasses image correction method based on factory calibration as described in claim 1, characterized in that, include: An image acquisition module includes a camera; the camera is integrated into the frame of the AI ​​glasses and is used to acquire raw images. The storage module includes a non-volatile memory for storing the inherent calibration parameters of the calibration unit; the inherent calibration parameters include a horizontal deviation angle and a vertical deviation angle. A calibration parameter reading module, which is connected to the storage module, is used to read the inherent calibration parameters of the AI ​​glasses; The calibration parameter calculation module, which is connected to the image acquisition module and the calibration parameter reading module, is used to calculate the geometric transformation parameters for correcting the original image based on the inherent calibration parameters and the camera's internal parameters. The image correction module, which connects the image acquisition module and the calibration parameter calculation module, is used to perform geometric transformation processing on the original image using geometric transformation parameters to generate a corrected image.

10. The AI ​​glasses image correction method based on factory calibration according to claim 9, characterized in that, Also includes: The field-of-view matching unit, which is integrated into the calibration parameter calculation module, is used to correct the geometric transformation parameters based on the difference between the standard human eye field of view and the camera field of view. An image filling module, which is connected to the image correction module, is used to perform edge filling or content-aware filling on the corrected image to generate a complete image.

Images