Camera defibrillation method and system

Abstract

The invention discloses a camera defibrillation method and system, and the method comprises the steps: collecting the operation parameters of a camera, determining a defibrillation scheme of the camera based on the operation parameters, and controlling a defibrillation assembly to execute the defibrillation scheme. According to the method, different defibrillation schemes can be formulated according to different operation conditions, and accurate control and correction of high-speed video stream image trembling are realized.

Description

Technical Field

[0001] The present invention relates to the technical field of computer vision detection, and in particular to a camera defibrillation method and system. Background Art

[0002] The packaging machine is a kind of equipment commonly used in the field of SMT packaging. It can package chips during high-speed operation. However, due to factors such as vibration and acceleration during the operation of the equipment, the high-speed video stream images captured by the camera usually have serious tremors, which affects the accuracy and reliability of subsequent image processing and analysis. Therefore, how to solve this problem has become one of the hot spots and difficulties in the current application field of SMT packaging technology.

[0003] Existing image stabilization technologies mainly include software solutions based on optical image stabilization, digital image stabilization, mechanical stabilizers, and image processing algorithms. However, for the image tremors under the high-speed video stream of the Jiufeng series micro packaging machine, the existing technologies have limitations and shortcomings such as low accuracy, poor adaptability, and high cost. Summary of the invention

[0004] The present invention provides a camera defibrillation method and system for solving the problem of jitter at the camera end of a chip mounter in the existing chip packaging process, which easily leads to inaccurate focus and blurred and unclear chip images.

[0005] In order to solve the above technical problems, the technical solution proposed by the present invention is:

[0006] A camera defibrillation method comprises the following steps:

[0007] The operating parameters of the camera are collected, a defibrillation scheme of the camera is determined based on the operating parameters, and a defibrillation component is controlled to execute the defibrillation scheme.

[0008] Preferably, determining the defibrillation scheme of the camera based on the operating parameters comprises the following steps:

[0009] A fuzzy function is used to determine the membership set of the operating parameters, and a preset defibrillation scheme corresponding to the membership set is used as the camera defibrillation scheme under the acceleration.

[0010] Preferably, the operating parameters include the vibration acceleration of the camera, the defibrillator assembly is an air-magnetic defibrillator, and the air-magnetic defibrillator is fixed on the camera; the air-magnetic defibrillator includes an air spring and a magnetic suspension isolator arranged in the air spring; the membership set includes a first set (0<ωd<5), a second set (5≤ωd≤20) and a third set (20<ωd<50), ωd is the magnitude of the camera vibration acceleration, and the value of the first set is smaller than the value of the second set, and the value of the second set is smaller than the value of the third set;

[0011] When the vibration acceleration belongs to the first set, the gas-magnetic defibrillator is not activated;

[0012] When the vibration acceleration belongs to the second set, determining the magnitude of the magnetic attraction current of the magnetic suspension isolator based on the vibration acceleration, and controlling the magnetic suspension isolator of the gas-magnetic defibrillator to execute the magnetic attraction current to achieve gas-magnetic anti-shake;

[0013] When the vibration acceleration belongs to the third set, the air spring of the air-magnetic defibrillator is activated to perform vibration isolation.

[0014] Preferably, the magnetic suspension vibration isolator includes a plurality of winding electromagnets, and an armature arranged between the plurality of winding electromagnets, which is used to fix a bracket assembly supporting the electromagnets and the armature.

[0015] Preferably, the bracket assembly includes a bracket bottom plate, a first bracket side plate, a second bracket side plate, a bracket middle plate, and a bracket top plate, the air spring is arranged in the space between the bracket bottom plate and the bracket top plate, the first bracket side plate, the second bracket side plate, and the bracket middle plate are arranged in the air spring, the bracket bottom plate and the bracket top plate extend laterally, the first bracket side plate, the second bracket side plate and the bracket middle plate both extend longitudinally, the first bracket side plate and the second bracket side plate are both fixed to the bracket bottom plate, the bracket middle plate is fixed to the bracket top plate and is arranged between the first bracket side plate and the second bracket side plate, and a laterally extending armature is fixed at a section away from the bracket top plate, and winding electromagnets are fixedly installed on the first bracket side plate and the second bracket side plate on both longitudinal sides of the armature.

[0016] Preferably, the method further comprises the following steps:

[0017] The image data collected by the camera is obtained, and it is determined whether the image data has an abnormality. If an abnormality exists, the abnormality category is determined, and a corresponding abnormality processing solution is executed on the image data to obtain denoised image data.

[0018] Preferably, the abnormal category includes inaccurate focus, and the abnormal processing scheme corresponding to the inaccurate focus is to denoise the image using a dynamic defibrillation algorithm based on the Doppler effect, and the dynamic defibrillation algorithm based on the Doppler effect performs the following steps:

[0019] Demodulate the original image data to baseband to obtain a point target model of the original image data;

[0020] Performing distance compression processing on the point target model to obtain a distance compression signal of the original image data;

[0021] Transforming the data at each range in the range compression signal into the range Doppler domain by azimuth FFT to obtain FFT transformed image data;

[0022] Compare the phase of the FFT transformed image data with the reference image, and determine the phase compensation value according to the comparison result;

[0023] Performing phase compensation on the FFT transformed image data based on the phase compensation value;

[0024] The phase-compensated FFT transformed image data is subjected to an inverse FFT transform to obtain image data after defibrillation.

[0025] Preferably, the abnormal category includes image afterimage, and the abnormal processing scheme corresponding to the image afterimage is deep network clearing processing, and the deep network clearing processing performs the following steps:

[0026] Construct the afterimage annealing operator H and its pseudo-inverse matrix of the input image data;

[0027] Constructing a diffusion network for removing residual images, and reconstructing the distribution space of input image data based on the diffusion network for removing residual images;

[0028] Get the content space of the input image data, and get the target annealed image based on the reconstructed distribution space and content space

[0029] Preferably, the abnormal category includes image blurring, and the abnormal processing scheme corresponding to the image blurring is to first denoise the image by a dynamic defibrillation algorithm based on the Doppler effect, and then perform deep network sharpening processing on the denoised image; determine whether the image after the deep network sharpening processing has a blurring abnormality, and if so, process the image after the deep network sharpening processing through the SRGAN deep sharpening network.

[0030] A computer system comprises a memory, a processor and a computer program stored in the memory and executable on the processor, wherein the steps of the above method are implemented when the processor executes the computer program.

[0031] The present invention has the following beneficial effects:

[0032] 1. The camera defibrillation method and system of the present invention collects the operating parameters of the camera, determines the defibrillation scheme of the camera based on the operating parameters, and controls the defibrillation component to execute the defibrillation scheme. The present invention can formulate different defibrillation schemes for different operating conditions, and realize accurate control and correction of high-speed video stream image tremors.

[0033] 2. In the preferred embodiment, the present invention uses gas magnetic suspension technology to reduce the vibration and tremor of the equipment during operation. By installing multiple precise magnetic suspension control systems on the equipment, the position, movement direction and speed of the equipment can be accurately controlled, thereby minimizing the vibration and tremor of the equipment. The application of this gas magnetic suspension technology can not only improve the clarity and stability of the image, but also ensure the performance and reliability of the equipment.

[0034] 3. In the preferred embodiment, the present invention proposes a dynamic defibrillation algorithm based on the Doppler effect and combines it with an image clarity model. The Doppler effect is used to compensate for the phase of the image to solve the problems of inaccurate focus and blurred images, and a deep network is used to make the image clearer.

[0035] 4. In the preferred embodiment, the present invention adopts a bidirectional defibrillator structure, uses an air-magnetic defibrillator assembly at the object end to reduce the impact of mechanical vibration on the camera, and uses a variety of image processing technologies at the viewing end to clarify the blurred image caused by shaking, thereby making the acquired image more stable and clear.

[0036] In addition to the above-described purposes, features and advantages, the present invention has other purposes, features and advantages. The present invention will be further described in detail below with reference to the accompanying drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0037] The drawings constituting a part of this application are used to provide a further understanding of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0038] Figure 1 is a partial structural diagram of a packaging machine according to an embodiment of the present invention;

[0039] Figure 2 is a hardware circuit diagram of a packaging machine according to an embodiment of the present invention;

[0040] Figure 3 is a structural diagram of a defibrillator assembly according to an embodiment of the present invention;

[0041] Figure 4 4 is a flow chart of a defibrillation method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0042] The embodiments of the present invention are described in detail below with reference to the accompanying drawings, but the present invention can be implemented in many different ways as defined and covered by the claims.

[0043] Embodiment 1:

[0044] The present invention provides a camera defibrillation method, which is applied to a packaging machine U6, such as Figure 1As shown, the packaging machine U6 includes a camera component, a computing drive component and a defibrillator component U4; the packaging machine U6 is used to package the electronic component U16; the camera component is installed on the packaging port of the packaging machine U6;

[0045] like Figure 2 As shown, the computing drive component includes a fuzzy controller U11, a CPU U12, a memory unit U17, an imaging drive module U19, an electromagnetic drive module U18, a video defibrillator module U21, an external interface module, and a bus U22; the CPU U12, the video defibrillator module U21, the external interface module, the imaging drive module U19, and the electromagnetic drive module U18 are connected via a bus U22; the external interface module U22 includes a monitoring interface and a sensor interface;

[0046] The camera assembly includes a CCD imaging chip U1, an imaging optical path U2, and an RGB camera lens U3; the imaging optical path U2 is equipped with a defibrillator assembly U4;

[0047] The CCD imaging chip U1, imaging optical path U2, RGB camera lens U3, and imaging driver module U19 are connected in sequence. The RGB camera lens U3 collects the package image of the electronic component U16, and transmits the package image to the CCD imaging chip U1 through the imaging optical path U2, and then sends it to the CPUU12 through the imaging driver module U19, and the CPUU12 sends it to the memory unit U17 for storage, and determines whether there is an abnormality in the package image. If there is an abnormality, the package image is sent to the visual end defibrillator module for processing.

[0048] like Figure 3 As shown, the defibrillator assembly U4 includes an acceleration sensor U7, an air spring U8, a winding electromagnet U9, an armature U10 and a bracket assembly; the bracket assembly includes a bracket bottom plate, a first bracket side plate, a second bracket side plate, a bracket middle plate, and a bracket top plate, the air spring U8 is arranged in the space between the bracket bottom plate and the bracket top plate, the first bracket side plate, the second bracket side plate, and the bracket middle plate are arranged in the air spring U8, the bracket bottom plate and the bracket top plate extend laterally, the first bracket side plate, the second bracket side plate and the bracket middle plate both extend longitudinally, the first bracket side plate and the second bracket side plate are both fixed on the bracket bottom plate, the bracket middle plate is fixed on the bracket top plate, and is arranged between the first bracket side plate and the second bracket side plate, and a section away from the bracket top plate is fixed with a laterally extending armature U10, and the first bracket side plate and the second bracket side plate on both longitudinal sides of the armature U10 are fixedly installed with a winding electromagnet U7; the acceleration sensor U7 is arranged on the outside of the top plate; the winding electromagnet U7 is electrically connected to the fuzzy control module U11.

[0049] The defibrillator assembly U4 in the present invention is an air-magnetic defibrillator assembly, and the defibrillation of the air-magnetic defibrillator assembly is achieved through the synergistic effect of magnetic levitation vibration isolation and air springs. The air spring fills the airbag with compressed air to provide flexible support for the camera equipment, and can adapt to working conditions under different loads by adjusting the inflation pressure. As a passive vibration isolator, it can effectively reduce the transmission of high-frequency dynamic vibrations. The electromagnetic force of the magnetic levitation isolator will be directly transmitted to the cover plate on the air spring, thereby acting on the equipment. For low-frequency vibrations, the low-frequency vibrations of the equipment can be offset by controlling the magnitude and direction of the electromagnetic force, ultimately achieving the isolation effect of the hybrid vibration isolator on high-frequency and low-frequency vibrations.

[0050] Assume that the rotating placement machine generates an exciting force where ω d is the excitation frequency, F d is the amplitude of the exciting force, j represents the imaginary unit, and t represents the time. Here we only consider the situation where the hybrid vibration isolator works normally (i.e., less than the maximum output electromagnetic force of the magnetic suspension vibration isolator), so the working principle of the gas-magnetic active and passive hybrid vibration isolator can be explained in three working conditions:

[0051] When the acceleration sensor input value is very small, that is, ω d The vibration isolator is in a static working state. The air spring provides the corresponding bearing capacity and the magnetic suspension isolator does not work.

[0052] When the acceleration sensor input value is small, that is, ω d Small, especially near the natural circular frequency of the air spring, the air spring amplifies the vibration, and the magnetic suspension isolator works to reduce or even suppress the resonance of the air spring.

[0053] According to the attached Figure 3 As shown in the figure, the magnetic suspension isolator consists of a bracket, a magnet, a differential coil, an armature, a displacement sensor and a force sensor. Since the magnetic attraction depends on the attraction of the magnetic force generated by the current in the differential coil on the armature, and based on the displacement parameters obtained by the displacement sensor, the electromagnetic force derived from the simplified electromagnetic circuit is as follows:

[0054]

[0055] When the parameters of the magnetic suspension isolator are determined, k is a constant, k = μ 0 N 2 A / 4, where μ 0 (H / m) is the magnetic permeability of air, N is the number of turns of the coil, A is the area of ​​the magnetic pole, x 0 is the air gap, x is the air gap change distance (i.e. the armature movement distance), i 1 ,i 2 are the currents of the upper and lower coils, pen d is the pressure generated by the current state of the air chamber, and S is the contact area between the air and the contact surface.

[0056] According to equation (1), the equivalent stiffness K and damping coefficient C of the vibration isolator are derived as follows:

[0057] From the ideal thermodynamic equation we can get:

[0058]

[0059] where p 0 is the initial pressure of the air chamber, T 0 is the temperature of the initial state (thermodynamic 7 temperature), and d is the initial gas chamber height.

[0060]

[0061]

[0062] in i 1 =i 10 ,i 2 =i 20 When F and C are linearized, they can be expanded by Taylor formula:

[0063]

[0064]

[0065]

[0066] Its linearized equivalent electromagnetic force, stiffness, and damping coefficient are shown in the above equations.

[0067] ③When the acceleration sensor input value is large, that is, ω d The air spring has a good vibration isolation effect, and the electromagnetic force required to be output by the magnetic suspension isolator is very small, which is close to a non-working state.

[0068] (2) The fuzzy control module U11 of the gas-magnetic defibrillator is composed of a fuzzy inference engine, a fuzzification module, a defuzzification module, a rule base and an input / output interface.

[0069] The working process of fuzzy control module U11 based on fuzzy control method is as follows:

[0070] Step 1: Input signal. The preprocessor acceleration signal and its rate detected by the acceleration sensor in the gas-magnetic defibrillator can be used as the input of the fuzzy controller.

[0071] Step 2: Fuzzification. Map the input acceleration sensor value a to the membership degree in the fuzzy set. A triangular or trapezoidal function can be used to define the shape and range of the fuzzy set. Based on expert experience, very small (NL) can be defined as 0-5Hz, small (SL) as 5-20Hz, and large (BL) as 20-50Hz.

[0072] Step 3: Fuzzy rule base. Create a series of fuzzy rules based on the characteristics and design requirements of the system:

[0073] 1. If d If it is NL, the magnetic de-shivering device does not work.

[0074] 2. If d If it belongs to SL, the magnetic attraction device is started, and the corresponding current feedback is performed according to the value of acceleration a to achieve a better gas-magnetic anti-shake effect.

[0075] 3. If d If it belongs to BL, the magnetic suction device is turned off and the air spring has a better anti-shake effect.

[0076] Step 4: Fuzzy reasoning. Based on the input fuzzy set and fuzzy rule base, the fuzzy output membership is calculated using the fuzzy reasoning method. This will result in a fuzzy output set that represents possible control actions.

[0077] Step 5: Defuzzification. Convert the fuzzy output set into a specific control action. The defuzzification method can be used to obtain the specific output electromagnetic force based on the defuzzification result.

[0078] Step 6: Feedback control. The current of the magnetic suspension isolator is used as the output of the fuzzy control module U11 to adjust the current of the electromagnet to achieve a better defibrillation effect. All these language variables are defined as NB (negative large), NS (negative small), ZE (zero), PS (positive small) and PB (positive large).

[0079] Based on the structure of the above packaging machine, Figure 4 As shown, the camera defibrillation method of the present invention has the following specific steps:

[0080] When the packaging machine starts packaging, if vibration occurs, the acceleration sensor U7 collects the corresponding acceleration of the vibration and transmits its acceleration signal to CPUU12 through bus U22.

[0081] CPUU12 determines whether to start the magnetic suspension defibrillator and the corresponding current size of the magnetic suspension defibrillator based on the acceleration signal size and the fuzzy control module U11 to perform object-side defibrillation.

[0082] Specifically, it is assumed that the fuzzy control module U11 includes three membership sets, namely, a first set, a second set, and a third set, and the value of the first set is smaller than the value of the second set, and the value of the second set is smaller than the value of the third set;

[0083] When the vibration acceleration belongs to the first set, the gas-magnetic defibrillator is not activated;

[0084] When the vibration acceleration belongs to the second set, determining the magnitude of the magnetic attraction current of the magnetic suspension isolator based on the vibration acceleration, and controlling the magnetic suspension isolator of the gas-magnetic defibrillator to execute the magnetic attraction current to achieve gas-magnetic anti-shake;

[0085] When the vibration acceleration belongs to the third set, the air spring of the air-magnetic defibrillator is activated to perform vibration isolation.

[0086] After the execution process starts, CPU U12 will perform anomaly detection and send an imaging control signal to the imaging drive unit U19, and control the RGB camera U3 through the bus U22 to collect the image of the electronic component U16 to be adjusted to the CCD imaging chip U1.

[0087] After being processed by the imaging drive unit U19, the digital image will be transmitted to the CPUU12 through the bus U22, and will also be output to the memory unit U17 through the bus U22.

[0088] CPUU12 continues to execute the abnormality detection task in the memory, determines the corresponding problems existing in the acquired RGB packaging image of the electronic component, and then performs Doppler effect defibrillation or depth clarity network clarity according to the corresponding problems.

[0089] Specifically, the anomaly detection task includes:

[0090] Perform edge detection on the contour curve of the electronic component in the package image to obtain the area where the electronic component is located;

[0091] Due to the high-speed rotation of the packaging machine, the video stream obtained by the camera may be out of focus. Therefore, a refocusing network is used to compensate the phase of the corresponding image through the Doppler effect dynamic defibrillation algorithm.

[0092] In the case of blurred images, the DDNM deep sharpening network is used to achieve image sharpening.

[0093] To address the problem of blurred images, we first use the Doppler effect dynamic defibrillation method to perform corresponding phase compensation and image adjustment on the image, and then use the SRGAN deep sharpening network to better explore the information and features in the data.

[0094] The specific contents of the dynamic defibrillation algorithm based on the Doppler effect are as follows:

[0095] The Doppler effect-based dynamic defibrillation algorithm estimates the motion trajectory and speed of an object or camera by analyzing image sequences or sensor data.

[0096] Based on the principle of the Doppler effect, the phase compensation amount for each pixel is calculated to correct the jitter and blur in the image. By applying the calculated phase compensation amount to each pixel in the image, the phase adjustment is achieved to minimize the impact of moving objects or cameras.

[0097] Perform azimuth fast Fourier transform to convert the data into azimuth Doppler domain. The specific steps are as follows:

[0098] The original image data is demodulated to baseband so that the middle of the range frequency is set to zero. The point target model after demodulation is:

[0099]

[0100] A 0 is an arbitrary complex constant, Γ is the range time, η is the close-range azimuth time, η c is the beam center deviation time; is the distance envelope (rectangular window function), that is:

[0101]

[0102] is the azimuth envelope (square function), that is

[0103]

[0104] f 0 is the image center frequency, K p is the distance chirp frequency, is the instantaneous slant range. The target azimuth time η is the echo S 0 The zero Doppler in (Γ,η) is the reference origin. For multiple targets, a common absolute time η should be given, such as the data capture start time η = 0. According to the translation property of Fourier transform, the phase term will be carried in the subsequent azimuth FFT.

[0105] Distance compression, let S 0 (f Γ ,η) is the formula S 0 The distance Fourier transform of (Γ,η) is is a matched filter, then the range compression output is:

[0106]

[0107] Among them, the compressed pulse network Pr [Γ] is the window function, W r (f τ ) is the fast Fourier transform of .

[0108] Directional angle fast Fourier transform, at low oblique angle, the beam points to the direction close to zero Doppler, and the distance equation can be approximated as a parabola:

[0109]

[0110] From the previous analysis (Formula 10), the range compression signal is:

[0111]

[0112] The azimuth phase modulation can be clearly seen from the second exponential term. Since the phase is η 2 The signal has a linear frequency modulation characteristic, and the frequency modulation is (approximation at low oblique viewing angle):

[0113]

[0114] Then, the data at each range is transformed into the range Doppler domain through azimuth FFT. For a given target, formula S rc The first exponential term in (Γ,η) is a constant, so only the second exponential term needs to be considered in the signal derivation in the range Doppler domain. Using the principle of stationary phase (POSP), the time-frequency relationship in azimuth is obtained as f η =-K a η, η=-f η / K a Substitute S rc (Γ,η), the signal of azimuth FFT is:

[0115]

[0116] W a (f η -f ηc ) is the direction diagram The frequency domain form of is the same in shape. The above formula contains two exponential terms. The first term is the inherent phase information of the target and has nothing to do with the image intensity. The second term is the frequency domain azimuth modulation with linear frequency modulation characteristics.

[0117] The phase of the image obtained by fast Fourier transform is compared with the reference image, and the phase compensation in the range Doppler domain can be defined as:

[0118]

[0119] The phase that needs to be compensated is:

[0120]

[0121] After the image deviation phase is compensated, the inverse fast Fourier transform is performed. The formula is as follows:

[0122]

[0123] Among them, p a is the amplitude of the azimuth impulse response, and outputs the image after adjusting the phase according to the calibration diagram.

[0124] Specifically, the specific contents of DDNM deep network clarification network are as follows:

[0125] First, the problem is mathematically modeled and the residual image annealing operator H is defined as the inverse matrix H of the residual image annealing operator. * The operator H simulates the image afterimage problem and Gaussian blur on the camera side during the shooting process. The afterimage and blurring process can be simulated by the formula y=Hx, where x represents a high-definition image without afterimage, and y represents the image taken by the camera. Therefore, the image y input in step (1) is subjected to the image inverse problem to try to inversely deduce the target annealing image. The removal of residual images is completed.

[0126]

[0127] Formula (7) gives the obtained de-ghosted image The constraints and the expression of the afterimage annealing operator H are shown in Figure 2. In H, Δt represents a time range from the current frame to the future frame, V t+τ The image optical flow field description with time delay τ, σ v , σ t , the motion time standard deviation parameter, respectively characterizes the influence of the object's moving speed on the afterimage effect, and the effect of the effect decaying over time.

[0128] In summary, the present invention adopts a bidirectional defibrillator structure, uses an air-magnetic defibrillator component at the object end to reduce the impact of mechanical vibration on the camera, and uses a variety of image processing technologies at the viewing end to clarify the blurred image caused by shaking, thereby making the acquired image more stable and clear.

[0129] In addition, a dynamic defibrillation algorithm based on the Doppler effect is proposed and combined with an image clarity model. The Doppler effect is used to compensate for the phase of the image to solve the problems of inaccurate focus and blurred image, and the image is made clearer through a deep network.

[0130] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A camera defibrillation method, characterized in that: The following steps are involved: The operating parameters of the camera are collected, a defibrillation scheme of the camera is determined based on the operating parameters, and a defibrillation component is controlled to execute the defibrillation scheme.

2. The camera defibrillation method according to claim 1, characterized in that: Determining a defibrillation scheme for the camera based on the operating parameters comprises the following steps: A fuzzy function is used to determine the membership set of the operating parameters, and a preset defibrillation scheme corresponding to the membership set is used as the camera defibrillation scheme under the acceleration.

3. The camera defibrillation method according to claim 2, characterized in that: The operating parameters include the vibration acceleration of the camera, the defibrillator assembly is an air-magnetic defibrillator, and the air-magnetic defibrillator is fixed on the camera; the air-magnetic defibrillator includes an air spring and a magnetic suspension isolator arranged in the air spring; the membership set includes a first set, a second set and a third set, ωd is the magnitude of the camera vibration acceleration, and the value of the first set is smaller than the value of the second set, and the value of the second set is smaller than the value of the third set; When the vibration acceleration belongs to the first set, the gas-magnetic defibrillator is not activated; When the vibration acceleration belongs to the second set, determining the magnitude of the magnetic attraction current of the magnetic suspension isolator based on the vibration acceleration, and controlling the magnetic suspension isolator of the gas-magnetic defibrillator to execute the magnetic attraction current to achieve gas-magnetic anti-shake; When the vibration acceleration belongs to the third set, the air spring of the air-magnetic defibrillator is activated to perform vibration isolation.

4. The camera defibrillation method according to claim 3, characterized in that: The first set is: 0<ωd<5, ωd is the camera vibration acceleration; the second set is: 5≤ωd≤20; the third set is 20<ωd<50; The magnetic suspension vibration isolator comprises a plurality of winding electromagnets and an armature arranged between the plurality of winding electromagnets, and is used for fixing a bracket assembly supporting the electromagnets and the armature.

5. The camera defibrillation method according to claim 4, characterized in that: The bracket assembly includes a bracket bottom plate, a first bracket side plate, a second bracket side plate, a bracket middle plate, and a bracket top plate. The air spring is arranged in the space between the bracket bottom plate and the bracket top plate. The first bracket side plate, the second bracket side plate, and the bracket middle plate are arranged in the air spring. The bracket bottom plate and the bracket top plate extend laterally. The first bracket side plate, the second bracket side plate, and the bracket middle plate both extend longitudinally. The first bracket side plate and the second bracket side plate are both fixed to the bracket bottom plate. The bracket middle plate is fixed to the bracket top plate and is arranged between the first bracket side plate and the second bracket side plate. A laterally extending armature is fixed at a section away from the bracket top plate. Winding electromagnets are fixedly installed on the first bracket side plate and the second bracket side plate on both longitudinal sides of the armature.

6. The camera defibrillation method according to any one of claims 1 to 5, characterized in that: The following steps are also included: The image data collected by the camera is obtained, and it is determined whether the image data has an abnormality. If an abnormality exists, the abnormality category is determined, and a corresponding abnormality processing solution is executed on the image data to obtain denoised image data.

7. The camera defibrillation method according to claim 6, characterized in that: The abnormality category includes inaccurate focus. The abnormality processing scheme corresponding to the inaccurate focus is to denoise the image using a dynamic defibrillation algorithm based on the Doppler effect. The dynamic defibrillation algorithm based on the Doppler effect performs the following steps: Demodulate the original image data to baseband to obtain a point target model of the original image data; Performing distance compression processing on the point target model to obtain a distance compression signal of the original image data; Transforming the data at each range in the range compression signal into the range Doppler domain by azimuth FFT to obtain FFT transformed image data; Compare the phase of the FFT transformed image data with the reference image, and determine the phase compensation value according to the comparison result; Performing phase compensation on the FFT transformed image data based on the phase compensation value; The phase-compensated FFT transformed image data is subjected to an inverse FFT transform to obtain image data after defibrillation.

8. The camera defibrillation method according to claim 6, characterized in that: The abnormal category includes image afterimage, and the abnormal processing scheme corresponding to the image afterimage is deep network clearing processing, and the deep network clearing processing performs the following steps: Construct the afterimage annealing operator H and its pseudo-inverse matrix of the input image data; Constructing a diffusion network for removing residual images, and reconstructing the distribution space of input image data based on the diffusion network for removing residual images; Get the content space of the input image data, and get the target annealed image based on the reconstructed distribution space and content space 9. The camera defibrillation method according to claim 6, characterized in that: The abnormality category includes image blurring. The abnormality processing scheme corresponding to the image blurring is to first denoise the image by a dynamic defibrillation algorithm based on the Doppler effect, and then perform deep network sharpening processing on the denoised image; determine whether the image after the deep network sharpening processing has a blurring abnormality, and if so, process the image after the deep network sharpening processing through the SRGAN deep sharpening network.

10. A computer system comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the method according to any one of claims 1 to 9 are implemented.

Images