Anti-shake system and method for fog-penetrating camera

Abstract

The invention mainly relates to the technical field of fog-penetrating cameras, and discloses a fog-penetrating camera anti-shake system and method, and the system comprises a driving module which adjusts the change of a light incident angle according to an instruction of a control circuit module; the control circuit module is connected with the inertial measurement module, the driving module and the electronic anti-shake module, and is used for processing data of the inertial measurement module and calculating compensation parameters; the inertial measurement module is used for measuring motion information of the equipment in multiple axial directions; and the electronic anti-shake module is used for carrying out electronic compensation processing on the image signal according to the compensation parameter. Therefore, the system can comprehensively and accurately sense the spatial motion of the camera, including linear motion and rotary motion, so that a more accurate anti-shake effect is realized.

Description

Technical Field

[0001] The present invention relates to the technical field of fog-penetrating cameras, and particularly to a fog-penetrating camera anti-shake system and method. Background Art

[0002] Fog-penetrating cameras have important application values in bad weather such as foggy days and sandstorms, as well as in moving shooting scenarios (such as vehicle-mounted monitoring, hand-held shooting, etc.). However, during actual use, the camera is easily affected by jitters caused by various factors. Whether it is the natural shaking of the human body during hand-held shooting, the bumpy vibration of the vehicle during vehicle-mounted installation, or the slight displacement of the device itself in a strong wind environment, these jitters will cause the imaging picture to be blurred and deformed, seriously affecting the quality and usability of the images captured by the fog-penetrating camera.

[0003] Traditional anti-shake technologies have certain limitations in the application scenarios of fog-penetrating cameras. For example, some optical anti-shake structures are complex and costly, making it difficult to be widely applied in mass-produced fog-penetrating camera products; some electronic anti-shake algorithms are not precise enough in dealing with the combined effects of light scattering interference and jitter in special weather conditions of fog-penetrating cameras, and cannot effectively eliminate the image degradation caused by jitter. Therefore, there is an urgent practical need to develop an efficient, low-cost, and precise anti-shake structure and algorithm suitable for fog-penetrating cameras. Summary of the Invention

[0004] In view of the above-mentioned problems in the prior art, the present invention is proposed.

[0005] The object of the present invention is to provide a fog-penetrating camera anti-shake system, aiming at the problems that existing fog-penetrating cameras have jitters and traditional anti-shake technologies have certain limitations in the application scenarios of fog-penetrating cameras.

[0006] To solve the above technical problems, the present invention provides the following technical solution: A fog-penetrating camera anti-shake system includes a driving module that adjusts the change of the light incident angle according to the instructions of the control circuit module;

[0007] The control circuit module is connected to the inertial measurement module, the driving module, and the electronic anti-shake module, processes the data of the inertial measurement module, and calculates compensation parameters;

[0008] The inertial measurement module is used to measure the motion information of the device in multiple axes;

[0009] The electronic anti-shake module performs electronic compensation processing on the image signal according to the compensation parameters.

[0010] As a preferred solution of the fog-penetrating camera anti-shake system of the present invention, wherein: The inertial measurement module is installed inside the fog-penetrating camera near the lens, and is used to accurately measure the acceleration and angular velocity information of the camera in three axes X, Y, and Z in real time.

[0011] As a preferred solution of the anti-shake system of the fog-penetrating camera of the present invention, the control circuit module includes a microcontroller, an interface circuit, a lens driving circuit, and an image sensor control circuit; and has fast data processing capabilities to meet real-time anti-shake requirements.

[0012] As a preferred solution of the anti-shake system of the fog-penetrating camera of the present invention, the driving module includes a lens housing, an electromagnetic driving component arranged on one side of the lens housing, a floating lens group arranged on one side of the electromagnetic driving component, an elastic support member arranged on one side of the floating lens group, a displacement detection sensor arranged on one side of the elastic support member, and a circuit interface arranged on one side of the displacement detection sensor.

[0013] As a preferred solution of the anti-shake system for a fog-penetrating camera of the present invention, the electromagnetic driving component includes a coil component surrounding one side of the floating lens group, and a permanent magnet fixedly mounted on the lens housing.

[0014] As a preferred solution of the anti-shake system of the fog-penetrating camera of the present invention, wherein: the coil assembly is composed of a plurality of independent coils corresponding to the driving in the X and Y directions respectively;

[0015] And the permanent magnet is in block or ring shape.

[0016] Another object of the present invention is to provide a method for anti-shake of a fog-penetrating camera, comprising the following steps:

[0017] Data collection, obtaining the motion information of the equipment;

[0018] Model building, using machine learning algorithms to train estimation models;

[0019] Compensation calculation, calculating compensation parameters based on the estimated model;

[0020] Perform adjustments, perform physical and electronic compensation according to compensation parameters, and optimize the compensation effect through feedback mechanisms.

[0021] As a preferred solution of the anti-shake method for a fog-penetrating camera of the present invention, the shake prediction model is established by adopting a neural network algorithm, and the model can quickly and accurately predict the blur degree and displacement direction of the image caused by camera shake based on the real-time motion data collected by the current inertial measurement unit.

[0022] As a preferred solution of the anti-shake method of the fog-penetrating camera of the present invention, it also includes image compensation processing, and performs actual electronic compensation processing operations on the image signal according to the calculated compensation parameters, including pixel displacement correction, difference compensation and gain adjustment, etc., to eliminate image blur and deformation caused by shaking.

[0023] As a preferred solution of the anti-shake method for the fog-penetrating camera of the present invention, it further includes initialization, in which the inertial measurement unit is calibrated to ensure the accuracy of its measurement data, and the pre-trained jitter prediction model and image quality evaluation algorithm are loaded into the storage unit of the anti-shake control circuit.

[0024] The beneficial effects of the anti-shake system for the fog-penetrating camera of the present invention are Description of the Drawings

[0025] In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings in the following description only relate to some embodiments of the present invention and do not limit the present invention. Among them:

[0026] Figure 1 It is a schematic diagram of the anti-shake system for the fog-penetrating camera of the present invention.

[0027] Figure 2 It is a sectional view of the structure entity of the drive module of the anti-shake system for the fog-penetrating camera of the present invention.

[0028] Figure 3 For the present invention Figure 2 An enlarged schematic view of the structure at A.

[0029] Figure 4 It is a flowchart of the anti-shake method for the fog-penetrating camera of the present invention. Detailed Embodiments

[0030] In order to enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below in conjunction with the specific embodiments and the drawings.

[0031] The terms used in the present invention are those general terms that are currently widely used in the art in consideration of the functions of the present invention, but these terms may change according to the intentions of those of ordinary skill in the art, precedents, or new technologies in the art. In addition, specific terms may be selected by the applicant, and in this case, their detailed meanings will be described in the detailed description of the present invention. Therefore, the terms used in the specification should not be understood as simple names, but based on the meanings of the terms and the overall description of the present invention.

[0032] Referring to Figure 1 , this embodiment provides an anti-shake system for a fog-penetrating camera, including a drive module 1 that adjusts the change in the light incident angle according to the instructions of the control circuit module;

[0033] A control circuit module, connected to the inertial measurement module, the drive module 1, and the electronic anti-shake module, processes the data of the inertial measurement module and calculates compensation parameters;

[0034] An inertial measurement module for measuring the motion information of the device in multiple axes;

[0035] The electronic image stabilization module performs electronic compensation processing on the image signal according to the compensation parameters.

[0036] In this embodiment, the inertial measurement module is installed inside the fog-penetrating camera near the lens, and is used to accurately measure the acceleration and angular velocity information of the camera on three axes X, Y, and Z in real time. This module adopts high-precision microelectromechanical sensor MEMS technology, can quickly respond to the minute movement changes of the camera, and transmits the measurement data to the subsequent processing module at a high rate. The inertial measurement module mainly includes an accelerometer and a gyroscope based on microelectromechanical sensor MEMS technology.

[0037] The control circuit module is connected to the inertial measurement module, the lens drive module of the camera, and the image sensor. Its main functions include preprocessing and analyzing the data transmitted from the inertial measurement module, and calculating the compensation displacement amount required for the lens and the electronic compensation parameters of the image sensor according to the preset algorithm model. This circuit uses an advanced digital signal processor DSP chip, which has powerful computing capabilities and fast data processing speeds to meet the requirements of real-time image stabilization. The control circuit module includes four parts, a microcontroller MCU, an IMU interface circuit, a lens drive circuit, and an image sensor control circuit.

[0038] As an alternative embodiment, the inertial measurement module is installed inside the fog-penetrating camera near the lens, and is used to accurately measure the acceleration and angular velocity information of the camera on three axes X, Y, and Z in real time.

[0039] As an alternative embodiment, the control circuit module includes a microcontroller, an interface circuit, a lens drive circuit, and an image sensor control circuit; and has fast data processing capabilities to meet the requirements of real-time image stabilization.

[0040] In an embodiment provided by the present application, the driving module 1 includes a lens housing 11, an electromagnetic driving component 12 disposed on one side of the lens housing 11, a floating lens group 13 disposed on one side of the electromagnetic driving component 12, an elastic support member 14 disposed on one side of the floating lens group 13, a displacement detection sensor 15 disposed on one side of the elastic support member 14, and a circuit interface 16 disposed on one side of the displacement detection sensor 15.

[0041] As an alternative embodiment, the electromagnetic driving component 12 includes a coil assembly 121 surrounding one side of the floating lens group 13, and a permanent magnet 122 fixedly installed on the lens housing 11.

[0042] As an alternative embodiment, the coil assembly 121 is composed of a plurality of independent coils, respectively corresponding to the driving in the X and Y directions;

[0043] Moreover, the permanent magnet 122 is in a block or ring shape.

[0044] In this embodiment, the overall shape of the optical image stabilization structure is cylindrical or approximately cylindrical, and it is installed at an appropriate position inside the lens. It is mainly designed around the optical axis of the lens to compensate for the change in the incident angle of light and ensure stable and clear imaging.

[0045] In this embodiment, the floating lens group 13 is usually composed of a set of 2-3 lenses. These lenses are thin and light, and are made of light and high-rigidity optical materials, such as special optical glass or optical plastics. Common ones like fluorite glass, which has excellent optical properties such as low dispersion and is relatively light, and polycarbonate optical plastics may be selected. The shapes of the lenses are various, such as biconvex, biconcave, plano-convex, plano-concave, etc., according to the optical design requirements. When combined together, they can effectively control the propagation and focusing of light. These lenses are not fixedly installed on the lens barrel, but can perform small displacement movements in a specific plane, that is, they can move relatively flexibly in the X horizontal and Y vertical directions to compensate for the change in the incident angle of light caused by camera shake.

[0046] In this embodiment, the electromagnetic drive assembly 12 is composed of a coil assembly 121 surrounding one side of the floating lens group 13 and a permanent magnet 122 fixedly installed on the lens housing 11. The coil assembly 121 surrounds the floating lens group and is usually composed of multiple independent coils, corresponding to the drives in the X and Y directions respectively.

[0047] There is one set or more sets of coils in the X direction, and there is a corresponding coil layout in the Y direction. The coils are generally wound with very thin enameled copper wires, and the number of turns is determined according to factors such as the required driving force. They are wound on a special magnetic skeleton, which can provide support for the coils and help guide the magnetic field direction.

[0048] The permanent magnet 122 corresponds to the coil assembly 121 and is fixedly installed on the housing or fixed bracket of the optical image stabilization structure, close to the coil. The permanent magnet 122 can provide a stable magnetic field. Common materials for the permanent magnet 122 include neodymium iron boron, etc. Because of its high magnetic field strength characteristics, it can generate a strong enough magnetic field to interact with the coil and generate an electromagnetic force to drive the floating lens group 13 to move. The permanent magnet 122 is generally in a block or ring shape, etc., and is reasonably placed according to the overall structure layout to ensure that the magnetic field is uniform and effectively acts on the coil.

[0049] A displacement detection sensor 15 is installed to monitor the actual displacement of the floating lens group in real time. Common ones include capacitive displacement sensors or Hall effect sensors, etc.

[0050] The floating lens group 13 is connected to the fixed part of the lens through an elastic support 14, and this structure plays two key roles. On the one hand, it allows the lens group to make small displacements under electromagnetic drive, and its elastic characteristics enable the lens group to quickly return to the initial position or reach a new equilibrium position when the driving force disappears or changes; on the other hand, it provides a certain support force for the lens group to ensure that the lenses maintain a stable relative position and attitude under normal circumstances and will not produce unnecessary offsets due to their own gravity or other external forces. The elastic support 14 is usually made of elastic materials such as special metal sheets, rubber elastomers, etc., and its shape may be a cantilever beam type, a spring sheet type, etc., which are selected and optimized according to specific design requirements and spatial layouts.

[0051] A circuit interface 16 is provided on one side or a suitable position of the lens anti-shake structure for connecting to an external anti-shake control circuit. This interface generally has multiple pins, which are respectively used for transmitting control signals such as drive current control signals from the anti-shake control circuit to adjust the magnitude and direction of the driving force of the electromagnetic drive component 12, receiving feedback signals (if there is a displacement detection sensor), feeding back the detected displacement data to the control circuit, and supplying power to the electromagnetic drive component 12 and the displacement detection sensor 15, etc.

[0052] In this embodiment, the inertial measurement module adopts high-precision microelectromechanical sensor MEMS technology, which can accurately measure the acceleration and angular velocity information of the camera in three axes X, Y, and Z in real time. This ensures that even under subtle motion changes, the system can quickly respond and provide accurate data. The inertial measurement module can not only quickly capture motion changes but also transmit data to the subsequent processing module at a high rate. This efficient communication ability ensures the real-time performance and reliability of the anti-shake system. The accelerometer and gyroscope are arranged perpendicular to each other in three axes and are integrally and compactly integrated on a chip or a substrate. Such a design enables the system to comprehensively and accurately sense the spatial motion of the camera, including linear movement and rotational movement, so as to achieve a more accurate anti-shake effect.

[0053] The control circuit module combines advanced machine learning algorithms such as neural networks, establishes a jitter prediction model based on the data provided by the inertial measurement module, and calculates the optimal compensation parameters. This method takes into account the influence of light refraction and scattering in a fog-penetrating environment and further improves the image quality.

[0054] Refer to Figure 2 , this embodiment provides a manufacturing method for a fog-penetrating camera anti-shake system, which is characterized in that it includes the following steps:

[0055] Data acquisition, obtaining the motion information of the device;

[0056] Model establishment, training a prediction model using machine learning algorithms;

[0057] Compensation calculation, calculating compensation parameters based on a prediction model;

[0058] Execute adjustment, perform physical and electronic compensation according to the compensation parameters, and optimize the compensation effect through a feedback mechanism.

[0059] In this embodiment, data acquisition, obtaining the motion information of the device. The inertial measurement unit IMU is installed inside the fog-penetrating camera near the lens, adopting high-precision microelectromechanical sensor MEMS technology. The IMU accurately measures the acceleration and angular velocity changes of the camera in three axes X, Y, and Z in real time, and collects these data at a high frequency, such as 1000Hz.

[0060] Model establishment, training a prediction model using machine learning algorithms. The system pre-collects a large amount of camera motion data and its corresponding image blur feature data under different environments. Using these data as the training set, a neural network model is trained through a deep learning framework such as TensorFlow or PyTorch. This model can quickly and accurately estimate the blur degree and displacement direction of the image caused by camera jitter based on the real-time motion data collected by the current inertial measurement unit. The jitter prediction model after sufficient training and verification is loaded into the storage unit of the anti-shake control circuit for calling during actual operation.

[0061] Compensation calculation, calculating compensation parameters based on the prediction model. When new motion data arrives, the control circuit module immediately inputs it into the prediction model for rapid prediction. According to the prediction result, calculate the physical compensation displacement amount that the lens needs to perform, that is, the moving distance of the floating lens group 13 in the X and Y directions, and the electronic compensation parameters required by the image sensor, such as pixel displacement correction, interpolation compensation, and gain adjustment. At the same time, considering the influence of light refraction and scattering in the fog-penetrating environment, the compensation amount is corrected and optimized.

[0062] Execute adjustment, perform physical and electronic compensation according to the compensation parameters, and optimize the compensation effect through a feedback mechanism. According to the instruction issued by the control circuit, after the coil assembly 121 is powered on, an electromagnetic force is generated under the magnetic field provided by the permanent magnet 122, causing the floating lens group 13 to undergo a small displacement in the required direction, thereby offsetting the change in the light incident angle caused by camera jitter.

[0063] As an alternative embodiment, a neural network algorithm is used to establish the jitter prediction model. This model can quickly and accurately estimate the blur degree and displacement direction of the image caused by camera jitter based on the real-time motion data collected by the current inertial measurement unit.

[0064] As an alternative embodiment, it further includes image compensation processing, which performs actual electronic compensation processing operations on the image signal according to the calculated compensation parameters, including pixel displacement correction, differential compensation, and gain adjustment, etc., to eliminate image blurring and distortion caused by jitter.

[0065] As an alternative embodiment, it further includes initialization, in which the inertial measurement unit is calibrated to ensure the accuracy of its measurement data, and the pre-trained jitter prediction model and image quality evaluation algorithm are loaded into the storage unit of the anti-shake control circuit.

[0066] As an alternative embodiment, when new motion data arrives, the control circuit module immediately inputs it into the prediction model for rapid prediction.

[0067] As an alternative embodiment, according to the prediction result, the physical compensation displacement amount that the lens needs to perform, that is, the moving distances of the floating lens group 13 in the X and Y directions, and the electronic compensation parameters required by the image sensor, such as pixel displacement correction, interpolation compensation, and gain adjustment, etc., are calculated. At the same time, considering the influence of light refraction and scattering in the fog-penetrating environment, the compensation amount is corrected and optimized.

[0068] As an alternative embodiment, according to the instruction issued by the control circuit, after the coil assembly 121 is energized, an electromagnetic force is generated under the magnetic field provided by the permanent magnet 122, causing the floating lens group 13 to undergo a small displacement in the required direction, thereby offsetting the change in the light incident angle caused by camera jitter. Ensure that the floating lens group 13 can quickly return to its original position or reach a new equilibrium position when the driving force disappears. The actual displacement of the lens group is monitored in real time to ensure that it moves as expected, and this information is fed back to the control circuit for further correction.

[0069] As an alternative embodiment, according to the compensation parameters received from the control circuit, the electronic anti-shake module performs operations such as pixel-level displacement correction and interpolation compensation on the image signal to ensure that the finally output image is clear and stable. Appropriately adjust the overall brightness and contrast of the image to optimize the visual effect.

[0070] As an alternative embodiment, the built-in image quality evaluation algorithm will automatically check whether the compensated image meets the predetermined quality standard. If it is found that the image quality does not meet the standard, the compensation parameters are fine-tuned according to the feedback result until a satisfactory imaging effect is obtained.

[0071] As an alternative embodiment, it includes pixel displacement correction, interpolation compensation, and gain adjustment, etc., to eliminate image blurring and distortion caused by jitter. These processing steps ensure that the clarity and stability of the image can be maintained even in a complex environment.

[0072] As an alternative embodiment, during the system assembly and initialization process, the inertial measurement unit is calibrated to ensure the accuracy of its measurement data. The pre-trained jitter prediction model and the image quality evaluation algorithm are loaded into the storage unit of the anti-shake control circuit to prepare for the normal operation of the system.

[0073] As an alternative embodiment, during the entire shooting process, the anti-shake system continuously repeats the above steps to adapt to the changing environmental conditions and camera poses, and always maintains the best anti-shake performance.

[0074] Finally, it should be noted that the methods and devices described in detail above are only embodiments, and those skilled in the art can modify these embodiments in different ways as long as they do not depart from the scope of the present invention.

Claims

1. A fog-penetrating camera anti-shake system, characterized by: include, A driving module (1) adjusts the incident angle of the light according to the instruction of the control circuit module; The control circuit module is connected to the inertial measurement module, the driving module (1) and the electronic anti-shake module, processes the data of the inertial measurement module and calculates compensation parameters; The inertial measurement module is used to measure the motion information of the device in multiple axes; The electronic anti-shake module performs electronic compensation processing on the image signal according to the compensation parameters.

2. The anti-shake system for a fog-penetrating camera as claimed in claim 1, characterized in that: The inertial measurement module is installed inside the fog-penetrating camera near the lens, and is used to accurately measure the acceleration and angular velocity information of the camera in three axes X, Y and Z in real time.

3. The anti-shake system for a fog-penetrating camera as claimed in claim 1, characterized in that: The control circuit module includes a microcontroller, an interface circuit, a lens driving circuit, and an image sensor control circuit; and has fast data processing capabilities to meet real-time anti-shake requirements.

4. The anti-shake system for a fog-penetrating camera as claimed in claim 3, characterized in that: The driving module (1) comprises a lens housing (11), an electromagnetic driving component (12) arranged on one side of the lens housing (11), a floating lens group (13) arranged on one side of the electromagnetic driving component (12), an elastic supporting member (14) arranged on one side of the floating lens group (13), a displacement detection sensor (15) arranged on one side of the elastic supporting member (14), and a circuit interface (16) arranged on one side of the displacement detection sensor (15).

5. The anti-shake system for a fog-penetrating camera as claimed in claim 4, characterized in that: The electromagnetic drive component (12) comprises a coil component (121) surrounding one side of the floating lens group (13), and a permanent magnet (122) fixedly mounted on the lens housing (11).

6. The anti-shake system for a fog-penetrating camera as claimed in claim 5, characterized in that: The coil assembly (121) is composed of a plurality of independent coils, corresponding to the driving in the X and Y directions respectively; Furthermore, the permanent magnet (122) is in a block or ring shape.

7. A method for anti-shake of a fog-penetrating camera, characterized in that: The following steps are involved: Data collection, obtaining the motion information of the equipment; Model building, using machine learning algorithms to train estimation models; Compensation calculation, calculating compensation parameters based on the estimated model; Perform adjustments, perform physical and electronic compensation according to compensation parameters, and optimize the compensation effect through feedback mechanisms.

8. The anti-shake method for a fog-penetrating camera according to claim 7, wherein: The jitter prediction model is established by using a neural network algorithm, and the model can quickly and accurately predict the blur degree and displacement direction of the image caused by camera jitter based on the real-time motion data collected by the current inertial measurement unit.

9. The anti-shake method for a fog-penetrating camera according to claim 8, wherein: It also includes image compensation processing, which performs actual electronic compensation processing operations on the image signal based on the calculated compensation parameters, including pixel displacement correction, interpolation compensation and gain adjustment, so as to eliminate image blur and deformation caused by jitter.

10. The anti-shake method for a fog-penetrating camera according to claim 9, characterized in that: It also includes initialization, in which the inertial measurement unit is calibrated to ensure the accuracy of its measurement data, and a pre-trained jitter prediction model and image quality assessment algorithm are loaded into the storage unit of the anti-shake control circuit.

Images