WO2023001036A1 - 摄像机 - Google Patents

摄像机 Download PDF

Info

Publication number
WO2023001036A1
WO2023001036A1 PCT/CN2022/105429 CN2022105429W WO2023001036A1 WO 2023001036 A1 WO2023001036 A1 WO 2023001036A1 CN 2022105429 W CN2022105429 W CN 2022105429W WO 2023001036 A1 WO2023001036 A1 WO 2023001036A1
Authority
WO
WIPO (PCT)
Prior art keywords
angle
shake
real
optical
time
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2022/105429
Other languages
English (en)
French (fr)
Inventor
王欢
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Hangzhou Hikvision Digital Technology Co Ltd
Original Assignee
Hangzhou Hikvision Digital Technology Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Hangzhou Hikvision Digital Technology Co Ltd filed Critical Hangzhou Hikvision Digital Technology Co Ltd
Publication of WO2023001036A1 publication Critical patent/WO2023001036A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules
    • H04N23/68—Control of cameras or camera modules for stable pick-up of the scene, e.g. compensating for camera body vibrations
    • H04N23/682—Vibration or motion blur correction
    • H04N23/685—Vibration or motion blur correction performed by mechanical compensation
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
    • H04N23/60—Control of cameras or camera modules
    • H04N23/68—Control of cameras or camera modules for stable pick-up of the scene, e.g. compensating for camera body vibrations
    • H04N23/682—Vibration or motion blur correction
    • H04N23/683—Vibration or motion blur correction performed by a processor, e.g. controlling the readout of an image memory

Definitions

  • the present application relates to anti-shake technology of cameras, in particular to a camera.
  • an embodiment of the present application provides a video camera, which helps reduce the risk of anti-shake failure.
  • the camera may include:
  • the shake sensor is used to continuously generate the real-time shake data of the camera at a first frequency, wherein the first frequency is not lower than twice the second frequency, and the second frequency is the imaging frequency of the photosensitive element of the camera frequency, and the real-time jitter data is used to determine the real-time spatial angle of the camera in the spatial coordinate system;
  • a processor component configured to update the weight value of optical anti-shake and the weight value of electronic anti-shake at the second frequency
  • the anti-shake driving component is used to implement the real-time spatial angle opposite to the real-time relative angle of the imaging coordinate system of the photosensitive element in the body coordinate system of the camera according to the weight value of the optical anti-shake Continuous angle correction to achieve optical image stabilization responsive to said real-time spatial angle;
  • angle correction is performed on the real-time relative angle according to the optical anti-shake weight value in the first imaging period of the photosensitive element, so that the real-time relative angle after angle correction is within the optical anti-shake capability range of the camera, so that the angle
  • the corrected real-time relative angle has a real-time angle difference relative to the real-time spatial angle
  • the processor component is further configured to determine the equivalent spatial angle of the first imaging cycle as the first equivalent spatial angle according to the real-time spatial angle in the first imaging cycle, and, according to the first imaging cycle
  • the electronic anti-shake weight value in the period and the first equivalent spatial angle determine the first electronic anti-shake compensation angle, so as to adjust the pixel array of the image obtained at the first refresh time according to the first electronic anti-shake compensation angle , the adjustment range of the pixel array is obtained, and the adjustment range of the pixel array is used to compensate the real-time angle difference to realize electronic anti-shake;
  • the first refresh moment represents the end moment of the first imaging period.
  • a control method for anti-shake including:
  • the real-time spatial angle in the first imaging cycle determine the equivalent spatial angle of the first imaging cycle as the first equivalent spatial angle
  • the obtained pixel array of the image obtains the adjustment range of the pixel array, and the adjustment range of the pixel array is used to compensate the real-time angle difference to realize electronic anti-shake.
  • a control device including: a processor; a memory for storing instructions executable by the processor; wherein the processor is configured to execute the steps of any one of the control methods described above .
  • a computer-readable storage medium is provided. Instructions are stored on the computer-readable storage medium. When the instructions are executed by a processor, the steps of any one of the control methods described above are implemented.
  • a computer program product including instructions is provided, and when it is run on a processor, it causes the processor to execute the steps of any one of the control methods described above.
  • any refresh period of the photosensitive element only a part of the real-time spatial angle of the camera within the scope of the optical image stabilization capability of the camera can be offset by the optical image stabilization, so as to help reduce the risk of optical image stabilization failure; Moreover, based on a part of the spatial angle offset by the optical image stabilization, the risk of image clarity being too low due to an excessively large spatial angle can also be reduced, so that at the refresh moment at the end of the imaging cycle, the electronic image stabilization pair has clarity Image processing of the image improved by optical image stabilization can compensate the impact of the remaining part of the real-time spatial angle of the camera on the image quality. Furthermore, through the dual-mode combination of electronic image stabilization and optical image stabilization, it helps to reduce the risk of image stabilization failure.
  • FIG. 1 is a schematic diagram of a hardware architecture of a camera in an embodiment of the present application
  • Fig. 2 is an exemplary schematic diagram of the real-time spatial angle determined by the camera shown in Fig. 1;
  • FIG. 3 is a schematic diagram of the principle of the cooperative anti-shake implemented by the camera shown in FIG. 1;
  • Fig. 4 is an exemplary effect diagram of electronic anti-shake compensation optical anti-shake adopted in cooperative anti-shake as shown in Fig. 3;
  • FIG. 5 is a schematic diagram of the principle of the weight update process applicable to the cooperative anti-shake as shown in FIG. 3;
  • Fig. 6 is an exemplary effect diagram of the optical excess range determined based on the speed limit during the weight update process as shown in Fig. 5;
  • Fig. 7 is an exemplary effect diagram of the optical excess range determined based on the pose limit during the weight update process as shown in Fig. 5;
  • Fig. 8 is an exemplary effect diagram of the electronic excess range determined by the associated shutter time during the weight update process as shown in Fig. 5;
  • Fig. 9 is a schematic flowchart of an exemplary anti-shake control method in another embodiment of the present application.
  • FIG. 1 is a schematic diagram of a hardware architecture of a camera in an embodiment of the present application.
  • the video camera may include a lens 10 and a body 20, wherein the lens 10 has an optical lens group 31, the body 20 has a photosensitive element 32, an anti-shake drive assembly 70 and a processor assembly 90, and the lens A shake sensor 50 may also be installed in the 10 or the body 20 .
  • the photosensitive element 32 can include CMOS (Complementary Metal Oxide Semiconductor, Complementary Metal Oxide Semiconductor) or CCD (Charge Coupled Device, Charge Coupled Device), used for exposure imaging in the imaging path of the optical mirror group 31, and refresh the output exposure imaging to obtain Image.
  • CMOS Complementary Metal Oxide Semiconductor, Complementary Metal Oxide Semiconductor
  • CCD Charge Coupled Device, Charge Coupled Device
  • the shake sensor 50 is used to continuously generate the real-time shake data of the camera with the first frequency f_gyro.
  • the shake sensor 50 may include a sensor such as a gyroscope for determining the real-time spatial angle of the camera, that is, the real-time spatial angle can be considered as the The real-time angle of the fuselage coordinate system in the space coordinate system.
  • the real-time shaking data continuously generated by the shaking sensor 50 with the first frequency f_gyro is used to determine the real-time spatial angle of the camera in the spatial coordinate system.
  • the first frequency f_gyro of the real-time shaking data generated by the shake sensor 50 is not lower than twice the second frequency f_img of the imaging of the photosensitive element 32 .
  • the real-time spatial angle includes an angle value sequence ⁇ A_i_1,...,A_i_n ⁇ determined according to at least two real-time shaking data, where i represents the ith imaging period T_img_i of the photosensitive element 32 , i is a positive integer greater than 0, and n is the number of real-time jitter data generated within one imaging cycle T_img_i, and n is greater than or equal to 2.
  • the second frequency f_img is also the frame rate of the video image generated by the camera.
  • Camera shake includes yaw shake in the horizontal swing direction, pitch shake in the tilt swing direction, and flip shake in the rotation direction around the optical axis of the lens.
  • the real-time shake data generated by the shake sensor 50 may be induced in response to at least one of yaw shake, pitch shake, and roll shake.
  • the real-time shake data continuously generated by the shake sensor 50 at the first frequency f_gyro may include responses to the camera's yaw shake, pitch shake, and roll shake.
  • At least one of the angular velocity or angular acceleration, and, through a preset algorithm, the real-time space angle ⁇ A_i_1, . . . , A_i_n ⁇ of the camera in the space coordinate system can be determined according to the angular velocity or angular acceleration.
  • the real-time space angle of the camera may include a yaw space angle reflecting yaw jitter, a pitch space angle reflecting pitch jitter, and a roll space angle reflecting roll jitter. That is, the general expression ⁇ A_i_1,...,A_i_n ⁇ used above to express the real-time spatial angles within one imaging cycle T_img_i can be embodied as: real-time yaw spatial angles ⁇ Yaw_i_1,...,Yaw_i_n ⁇ , or real-time elevation Pitch space angles ⁇ Pitch_i_1, ..., Pitch_i_n ⁇ , or real-time flip space angles ⁇ Roll_i_1, ..., Roll_i_n ⁇ .
  • FIG. 2 is an exemplary schematic diagram of the real-time spatial angle determined by the camera shown in FIG. 1 .
  • the change state of the real-time space angle ⁇ A_i_1,...,A_i_n ⁇ in the corresponding shake direction the real-time space angle ⁇ A_i_1 , ..., A_i_n ⁇ in Fig. 2 take the frequency period Tgyro of the first frequency f_gyro as the interval discrete distribution, and Fig.
  • the real-time spatial angle ⁇ A_i_1, ..., A_i_n ⁇ determined according to the real-time shaking data generated by the shaking sensor 50 is an angle value relative to a pre-calibrated spatial reference angle A_abs_ref, which is defined in space Demarcated in the coordinate system, for example, the space reference angle A_abs_ref may represent an expected angle of the optical axis of the lens in the deployment space of the camera.
  • the real-time yaw spatial angle ⁇ Yaw_i_1,..., Yaw_i_n ⁇ is the angle value relative to the pre-calibrated spatial yaw reference angle Yaw_abs_ref
  • the real-time pitch spatial angle ⁇ Pitch_i_1,..., Pitch_i_n ⁇ is relative to the pre-calibrated
  • the angle values of the spatial pitch reference angle Pitch_abs_ref and the real-time roll spatial angles ⁇ Roll_i_1, . . . , Roll_i_n ⁇ are angle values relative to the pre-calibrated spatial roll reference angle Roll_abs_ref.
  • each imaging period T_img_i at least for yaw shake and pitch shake, the cooperative image stabilization of optical image stabilization and electronic image stabilization can be implemented; while the flip jitter can completely rely on electronic image stabilization, or, Synergistic image stabilization of optical image stabilization and electronic image stabilization can be implemented against rollover shake.
  • the anti-shake driving assembly 70 can be used to realize optical anti-shake, specifically, the anti-shake driving assembly 70 can be used for the real-time relative angle of the imaging coordinate system of the photosensitive element 32 in the body coordinate system of the camera, and the real-time spatial angle ⁇ A_i_1, ..., A_i_n ⁇ reverse continuous angle correction, that is, the anti-shake driving component 70 can implement continuous angle correction, and the direction of the angle correction is opposite to the real-time space angle, so as to realize the response to the real-time space angle ⁇ A_i_1 ,..., A_i_n ⁇ optical image stabilization.
  • the real-time relative angle of the imaging coordinate system of the photosensitive element 32 can be determined by the relative pose between the optical mirror group 31 and the photosensitive element 32, therefore, the anti-shake drive assembly 70 can The relative pose between the elements 32 realizes the continuous angle correction of the imaging coordinate system.
  • the anti-shake driving assembly 70 may include an OIS driver for driving the relative pose adjustment between the optical lens group 31 and the photosensitive element 32 .
  • the continuous angle correction implemented by the anti-shake driving assembly 70 may be triggered in response to at least one of yaw shake, pitch shake and roll shake.
  • the relative pose adjusted by the anti-shake drive assembly 70 may include the adjustment of the relative translation position between the optical lens group 31 and the photosensitive element 32 .
  • the anti-shake drive assembly 70 can adjust the optical mirror group 31 and the photosensitive element 32.
  • the relative horizontal position or the relative vertical position between them so that the imaging coordinate system swings to a specified horizontal relative angle or a specified vertical relative angle relative to the camera in the direction opposite to the yaw shake or pitch shake of the camera, so as to Alleviate the fluctuation of the imaging coordinate system following the camera's yaw shake or pitch shake (that is, the yaw shake or pitch shake of the fuselage coordinate system in the space coordinate system), and promote the space of the imaging coordinate system in the space coordinate system
  • the angle remains steady.
  • the optical anti-shake achieved through the translation of the optical mirror group 31 may also be called lens anti-shake or optical mirror group displacement anti-shake.
  • the anti-shake drive assembly 70 can also adjust the optical mirror group 31 and The relative horizontal position or relative vertical position between the photosensitive elements 32 is used to alleviate the fluctuation of the imaging coordinate system following the yaw shake or pitch shake of the camera.
  • the optical anti-shake achieved by the movement of the photosensitive element 32 may also be referred to as body anti-shake or photosensitive element displacement anti-shake.
  • the relative pose adjusted by the anti-shake drive assembly 70 can include the relative translation angle between the optical mirror group 31 and the photosensitive element 32 adjustment.
  • the photosensitive element 32 is rotatably installed in the body 20 of the camera.
  • the anti-shake drive assembly 70 can adjust the distance between the optical mirror group 31 and the photosensitive element 32.
  • the relative translation angle so that the imaging coordinate system is flipped to the specified relative flip angle relative to the camera in the direction opposite to the camera’s flip jitter, so as to alleviate the imaging coordinate system following the camera’s flip jitter (that is, the fuselage coordinate system is in Flip jitter in the spatial coordinate system) fluctuations, and make the spatial angle of the imaging coordinate system in the spatial coordinate system remain stable.
  • Processor component 90 may be used to implement electronic image stabilization. Specifically, the processor component 90 can be used to perform image processing on the image output by the photosensitive element 32 at the refresh moment at the end of any imaging period T_img_i, so as to realize electronic anti-shake through pixel array adjustment generated by image processing.
  • processor component 90 may include an image processor for performing image processing on images. That is to say, the processor component 90 can adjust the pixel array of the image output by the photosensitive element 32 at the refresh moment at the end of any imaging period T_img_i to obtain the adjustment range of the pixel array, based on which the electronic anti-shake can be realized.
  • the adjustment of the pixel array by the processor component 90 through image processing may be triggered in response to at least one of yaw jitter, pitch jitter and flip jitter. That is, for the electronic anti-shake of yaw shake and pitch shake, the image output by the photosensitive element 32 can be shifted by the pixel array through the processor component 90; The output image undergoes pixel array translation.
  • the electronic image stabilization performed at the refresh moment of any imaging period T_img_i can use the single-frame equivalent spatial angle A_eq_i associated with the real-time spatial angle ⁇ A_i_1,...,A_i_n ⁇ within this imaging period T_img_i, namely , the single-frame equivalent yaw space angle Yaw_eq_i associated with the real-time yaw space angle ⁇ Yaw_i_1,..., Yaw_i_n ⁇ , the single-frame equivalent pitch space angle Pitch_eq_i associated with the real-time pitch space angle ⁇ Pitch_i_1,..., Pitch_i_n ⁇ , and, the single-frame
  • the single-frame equivalent spatial angle A_eq_i is associated with the real-time spatial angle ⁇ A_i_1,...,A_i_n ⁇ within the imaging period T_img_i, and the imaging period T_img_i represents the duration of an image frame, that is, the single-frame equivalent spatial angle A_eq_i is corresponding to the image frame, so it can be called a single-frame equivalent space angle, and it can also be called an equivalent space angle in this application.
  • the processor component 90 can sample the real-time spatial angles ⁇ A_i_1, ..., A_i_n ⁇ in any imaging period T_img_i to obtain the single-frame equivalent spatial angle A_eq_i of the imaging period T_img_i, that is, the single-frame equivalent spatial angle A_eq_i may also be called the sampling space angle.
  • the processor component 90 can sample the real-time spatial angles ⁇ A_i_1, ..., A_i_n ⁇ in any imaging period T_img_i to obtain the peak angle A_i_max (the absolute value of which is the largest), or the mean angle A_i_avg (the average value of the absolute value of the angle ), or the median angle A_i_med (the median value of the absolute value of the angle), and, the peak angle A_i_max, or the mean angle A_i_avg, or the median
  • the angle A_i_med is determined as the single-frame equivalent spatial angle A_eq_i within the imaging period T_img_i, namely:
  • the single-frame equivalent spatial angle A_eq_i represents the single-frame equivalent yaw spatial angle Yaw_eq_i
  • the median yaw angle Yaw_i_med can be determined as the single-frame equivalent yaw spatial angle Yaw_eq_i;
  • the single-frame equivalent space angle A_eq_i represents the single-frame equivalent pitch space angle Pitch_eq_i
  • the pitch peak angle Pitch_i_max, or the pitch average angle Pitch_i_avg, or The pitch median angle Pitch_i_med can be determined as the single-frame equivalent pitch space angle Pitch_eq_i;
  • the single-frame equivalent spatial angle A_eq_i represents the single-frame equivalent flipped spatial angle Roll_eq_i
  • the flipped peak angle Roll_i_max, or the flipped average angle Roll_i_avg, or the flipped median angle in the real-time flipped spatial angle ⁇ Roll_i_1,..., Roll_i_n ⁇ Roll_i_med can be determined as the single-frame equivalent flip space angle Roll_eq_i.
  • the peak angle A_i_max can reflect the maximum shaking amplitude that has occurred in the imaging period T_img_i, therefore, preferably, the peak angle A_i_max can be determined as the single-frame equivalent spatial angle A_eq_i, so as to help the electronic image stabilization to make the pixels of the image
  • the magnitude of modulation produced by the array is sufficient to mitigate or even eliminate the effect of the maximum dither magnitude on image quality.
  • the processor component 90 is also used to realize the cooperative control between the optical image stabilization and the electronic image stabilization.
  • the cooperative control may be undertaken by the image processor in the processor component 90 , or may be undertaken by another processor further included in the processor component 90 that is independent of the image processor.
  • the optical image stabilization weight Wois For the coordinated control of optical image stabilization and electronic image stabilization, in this embodiment, two parameters for implementing cooperative image stabilization are provided for the camera, namely, the optical image stabilization weight Wois and the electronic image stabilization weight Weis.
  • the weight value Wois_i of the optical anti-shake weight Wois and the weight value Weis_i of the electronic anti-shake weight Weis can be updated and adjusted synchronously with the image refresh of the photosensitive element 32 .
  • the weight value Wois_i of the optical image stabilization weight Wois may be called an optical image stabilization weight value
  • the weight value Weis_i of the electronic image stabilization weight Weis may be called an electronic image stabilization weight value.
  • the processor component 90 can be used to update the weight value Wois_i of the optical anti-shake weight Wois for realizing cooperative anti-shake and the weight value Weis_i of the electronic anti-shake weight Weis with the second frequency f_img, and the update basis can be a single imaging cycle T_img_i
  • the update of the weight value Wois_i of the optical anti-shake weight Wois by the processor component 90 and the weight value Weis_i of the electronic anti-shake weight Weis may be in response to at least one of yaw shake, pitch shake and roll shake. is raised, in this case:
  • the optical anti-shake weight Wois and the electronic anti-shake weight Weis can represent the optical yaw anti-shake weight Wois_yaw and the electronic yaw anti-shake weight Weis_yaw corresponding to the real-time yaw space angle ⁇ Yaw_i_1,..., Yaw_i_n ⁇ ;
  • optical anti-shake weight Wois and the electronic anti-shake weight Weis can also represent the optical anti-shake weight Wois_pitch and the electronic anti-shake weight Weis_pitch corresponding to the real-time pitch space angle ⁇ Pitch_i_1,...,Pitch_i_n ⁇ ;
  • the optical image stabilization weight Wois and the electronic image stabilization weight Weis can also represent the corresponding real-time flipping space angle ⁇ Roll_i_1 ,..., Roll_i_n ⁇ 's optical flip anti-shake weight Wois_roll and electronic flip anti-shake weight Weis_roll.
  • the processor component 90 updates the weight values of the optical yaw anti-shake weight Wois_yaw and the electronic yaw anti-shake weight Weis_yaw, updates the weights of the optical pitch anti-shake weight Wois_pitch and the electronic pitch anti-shake weight Weis_pitch, and updates the optical flip
  • the weight value update of the anti-shake weight Wois_roll and the electronic flip anti-shake weight Weis_roll can be independent of each other.
  • FIG. 3 is a schematic diagram of the principle of cooperative anti-shake implemented by the camera shown in FIG. 1 . See Figure 3:
  • the real-time spatial angle ⁇ A_i_1, . . . , A_i_n ⁇ is determined according to the real-time shake data generated by the shake sensor 50 .
  • the anti-shake driving component 70 can be specifically used for, in any imaging period T_img_i, according to the weight value Wois_i of the optical anti-shake weight Wois in the imaging period T_img_i, the real-time relative angle of the imaging coordinate system of the photosensitive element 32 relative to the camera, A continuous angle correction inverse to the real-time spatial angles ⁇ A_i_1,...,A_i_n ⁇ is implemented to achieve optical image stabilization responsive to the real-time spatial angles ⁇ A_i_1,...,A_i_n ⁇ .
  • the real-time relative angle of the imaging coordinate system relative to the camera is corrected to ⁇ Aosi_i_1, . . . , Aois_i_n ⁇ .
  • Aois_i_j A_i_j ⁇ Wois_i/(Wois_i+Weis_i)
  • j is the time-based sorting position in a sequence of real-time shaking data or real-time spatial angles within one imaging period T_img_i
  • j is greater than or equal to 1 and less than or equal to the aforementioned n.
  • the corrected real-time relative angle Aois_i_j refers to the imaging coordinate system relative to the space reference angle A_abs_ref caused by the real-time space angle ⁇ A_i_1,...,A_i_n ⁇ in order to offset the camera shake (that is, the space reference Angle), so that the angle value of the imaging coordinate system relative to the pre-marked relative reference angle A_rlv_ref, the relative reference angle A_rlv_ref can be demarcated in the body coordinate system of the camera, for example, the relative reference angle A_rlv_ref can represent The assembly angle of the optical axis of the lens in the body coordinate system, and the azimuth angle of the relative reference angle A_rlv_ref in the space coordinate system changes in response to the real-time space angle ⁇ A_i_1, ..., A_i_n ⁇ .
  • the electronic anti-shake compensation angle Aeis_i A_eq_i ⁇ Weis_i/(Wois_i+Weis_i) on which the pixel array adjustment of the electronic image stabilization is based.
  • the processor component 90 can be specifically configured to determine the single-frame equivalent spatial angle A_eq_i of the imaging period T_img_i according to the real-time spatial angle ⁇ A_i_1, ..., A_i_n ⁇ in any imaging period T_img_i, and, according to the electronic image stabilization
  • the weight value Weis_i of the weight Weis in the imaging period T_img_i and the single-frame equivalent spatial angle A_eq_i of the imaging period T_img_i determine the first electronic anti-shake compensation angle Aeis_i, so that the first electronic anti-shake compensation angle Aeis_i
  • the image 30 obtained at the refresh moment at the end of the imaging period T_img_i is subjected to image processing, so that the image is generated to realize the adjustment of the pixel array for electronic image stabilization.
  • Yaw_i_n ⁇ causes the yaw deviation of the imaging coordinate system relative to the space yaw reference angle Yaw_abs_ref, and makes the imaging coordinate system relative to the relative yaw reference angle Yaw_rlv_ref (such as the yaw direction of the lens optical axis in the fuselage coordinate system The angle value of the assembly angle above);
  • Wois_yaw_i represents the optical yaw anti-shake weight corresponding to the real-time yaw space angle Yaw_i_j
  • Weis_yaw_i represents the electronic yaw anti-shake weight corresponding to the real-time yaw space angle Yaw_i_j.
  • Pitch_i_n ⁇ causes the pitch deviation of the imaging coordinate system relative to the space pitch reference angle Pitch_abs_ref, and makes the imaging coordinate system relative to the relative pitch reference angle Pitch_rlv_ref (such as the pitch direction of the lens optical axis in the fuselage coordinate system The angle value of the assembly angle above);
  • Wois_pitch_i represents the optical pitch anti-shake weight corresponding to the real-time pitch space angle Pitch_i_j
  • Weis_pitch_i represents the electronic pitch anti-shake weight corresponding to the real-time pitch space angle Pitch_i_j.
  • the update of the weight value Wois_i of the optical image stabilization weight Wois by the processor component 90 and the weight value Weis_i of the electronic image stabilization weight Weis is based on the ability to correct the real-time relative angle of the optical image stabilization to the imaging coordinate system , and the consideration of the tolerance of electronic image stabilization to image clarity, this is because:
  • Optical anti-shake is limited by the physical adjustment performance of the relative pose between the optical lens group 31 and the photosensitive element 32. If the imaging coordinate system is corrected to be sufficient to offset the real-time relative angle of the real-time space angle ⁇ A_i_1,...,A_i_n ⁇ If the physical adjustment performance of the optical mirror group 31 and the photosensitive element 32 exceeds the physical adjustment performance of the relative pose between the optical lens group 31 and the photosensitive element 32, it is easy to cause the failure of the optical anti-shake.
  • Electronic anti-shake depends on the clarity of the image obtained by the photosensitive element 32. If the imaging coordinate system during the imaging cycle (especially during the exposure period in the imaging cycle) produces a large vibration due to an excessive real-time space angle, and causes the image to be clear If the brightness is reduced to a level that is not enough to be improved by image processing (such as blurring), it is easy to cause the failure of electronic anti-shake.
  • the camera’s real-time spatial angle ⁇ A_i_1,...,A_i_n ⁇ can only be part of the camera’s optical image stabilization capability (that is, the corrected real-time relative angle ⁇ Aois_i_1,..., Aois_i_n ⁇ reverse equivalent part) is offset by optical image stabilization to help reduce the risk of optical image stabilization failure; and, based on a part of the spatial angle offset by optical image stabilization, image clarity can also be reduced due to excessive spatial angles However, the risk is too low.
  • electronic image stabilization is used to perform image processing on the image whose definition has been improved by optical image stabilization, which can compensate the real-time spatial angle of the camera ⁇ A_i_1,..., The influence of the remaining part of A_i_n ⁇ (that is, the difference between the absolute value of A_i_j and the absolute value of Aois_i_j) on the image quality. Furthermore, through the dual-mode combination of electronic image stabilization and optical image stabilization, it helps to reduce the risk of image stabilization failure.
  • the initial value Wois_0 of the optical anti-shake weight Wois may be 1, and the initial value Weis_0 of the electronic anti-shake weight Weis may be 0. That is, the processor component 90 can be further configured to respond to the completion of the power-on of the camera, initially assigning the optical image stabilization weight Wois to 1, and initially assigning the electronic image stabilization weight Weis to 0, so as to realize the optical image stabilization priority synergy principle. That is to say, the weight value of the optical anti-shake may be initially assigned a value of 1, and the weight value of the electronic anti-shake may be initially assigned a value of 0.
  • the initial value Wois_yaw_0 of the optical yaw anti-shake weight Wois_yaw can be 1
  • the initial value Weis_yaw_0 of the electronic yaw anti-shake weight Weis_yaw can be 0
  • the initial value Wois_pitch_0 of the optical pitch anti-shake weight Wois_pitch can be 1
  • the electronic pitch anti-shake The initial value Weis_pitch_0 of the weight Weis_pitch can be 0
  • the initial value Wois_roll_0 of the optical anti-shake weight Wois_roll can be 1;
  • any imaging period T_img_i of the photosensitive element 32 is regarded as the first imaging period, and it is assumed that:
  • the weight value Wois_i of the optical anti-shake weight Wois in the first imaging period T_img_i of the photosensitive element 32 is intended to limit the real-time relative angle ⁇ Aois_i_1,...,Aois_i_n ⁇ within the optical anti-shake capability range of the camera, so that the first imaging period
  • the real-time relative angles ⁇ Aois_i_1, ..., Aois_i_n ⁇ can be corrected according to the weight value of optical image stabilization in the first imaging cycle, and then the real-time relative angles ⁇ Aois_i_1, ..., Aois_i_n ⁇ after angle correction can be
  • the camera’s optical anti-shake capability is within the range, and the angle-corrected real-time relative angle ⁇ Aois_i_1,...,Aois_i_n ⁇ has a real-time angle difference ⁇ A_i_1,..., ⁇ A_i_n ⁇ relative to the real-time space angle ⁇ A_i_1,...,A_i_n ⁇ .
  • FIG. 4 is an exemplary effect diagram of the electronic anti-shake compensation optical anti-shake adopted in the coordinated anti-shake as shown in FIG. 3 .
  • the fitting curve W51 represents the ideal value of the real-time relative angle ⁇ Aois_i_1, ..., Aois_i_n ⁇ when the optical yaw anti-shake weight Wois_yaw takes the initial value 1
  • the fitting curve W52 represents the real value of the real-time relative angle ⁇ Aois_i_1,...,Aois_i_n ⁇ actually occurring in the first imaging period T_img_i
  • the fitting curve W51 can be called the ideal real-time relative angle ⁇ Aois_i_1,...,Aois_i_n ⁇
  • the fitting curve, the fitting curve W52 can be called the real fitting curve of the real-time relative angle ⁇ Aois_i_1,...,Aois_i_n ⁇
  • the real-time angle difference ⁇ A_i_1, . . . , ⁇ A_i_n ⁇ can be equivalently represented by the deviation between the ideal fitting curve W51 and the real fitting curve W52.
  • the pixel arrays of ⁇ A_i_1, ..., ⁇ A_i_n ⁇ are adjusted to achieve electronic image
  • the cooperative image stabilization of the optical image stabilization and the electronic image stabilization can improve the image stabilization in the camera.
  • the quality of the image captured during the dithering period compared to the case of relying entirely on optical image stabilization to continuously implement full corrections to the relative coordinates of the imaging coordinate system, collaborative image stabilization can alleviate the problem when the real-time spatial angle ⁇ A_i_1,...,A_i_n ⁇ exceeds the range of optical image stabilization capabilities.
  • collaborative image stabilization can reduce the real-time spatial angle ⁇ A_i_1,...,A_i_n ⁇ enough to cause image blurring. Risk of electronic image stabilization failing due to image blur.
  • FIG. 5 is a schematic diagram of the principle of a weight update process applicable to the cooperative anti-shake shown in FIG. 3 . Please refer to FIG. 5, assuming that the previous imaging period T_img_i-1 of the first imaging period T_img_i is regarded as the second imaging period, then:
  • the processor component 90 may be further configured to determine the optical anti-shake weight according to the real-time spatial angle ⁇ A_i-1_1, ..., A_i-1_n ⁇ and the second electronic anti-shake compensation angle Aeis_i-1 in the second imaging period T_img_i-1
  • the weight value Weis_i of Wois in the first imaging period T_img_i and the weight value Weis_i of the electronic anti-shake weight Weis in the first imaging period T_img_i to promote:
  • the real-time relative angle ⁇ A_i_1, ..., A_i_n ⁇ is intended to be limited to converge within a preset angle correction limit range, wherein the angle correction limit range is determined according to the optical image stabilization capability range; as well as,
  • the sharpness of the image 30 obtained at the first refresh moment of the first imaging period T_img_i is not lower than the preset sharpness threshold, wherein the sharpness threshold is the lowest image preset for avoiding electronic image stabilization failure Clarity is ok.
  • the weight value of the optical image stabilization in the first imaging period can be determined, and the real-time relative angle in the first imaging period can be corrected based on the weight value of the optical stabilization.
  • the real-time relative angle after angle correction The angle converges within the preset angle correction limits.
  • the processor component 90 can be further used for:
  • the real-time spatial angle ⁇ A_i-1_1, ..., A_i-1_n ⁇ in the second imaging period T_img_i-1 determine the second single-frame equivalent spatial angle A_eq_i-1 in the second imaging period T_img_i-1, and , determining the magnitude by which the second single-frame equivalent spatial angle A_eq_i-1 exceeds the angle correction limit range, as the optical excess magnitude Aois_ex_i-1.
  • the optical excess amplitude Aois_ex_i-1 may also be referred to as the first amplitude.
  • the second single-frame equivalent spatial angle A_eq_i-1 is also the single-frame equivalent spatial angle associated with the second imaging period T_img_i-1.
  • the optical excess amplitude Aois_ex_i-1 produces a weight change trend that increases the weight value of the electronic anti-shake weight Weis, that is, the optical excess amplitude Aois_ex_i-1 promotes the weight of the electronic anti-shake weight Weis in the first imaging period T_img_i
  • the value Weis_i is greater than the weight value Weis_i-1 in the second imaging period T_img_i-1. That is to say, the increase of the first amplitude can cause the electronic anti-shake weight value to have an increasing change trend.
  • the angular deviation of the electronic anti-shake compensation angle frame deviation ⁇ Aeis_i-1
  • the electronic excess range Aeis_ex_i-1 may also be referred to as a second range.
  • the electronic excess range Aeis_ex_i-1 may cause a decrease in the definition of the image 30, and the electronic excess range Aeis_ex_i-1 produces a weight change trend that increases the weight value of the optical anti-shake weight Wois, that is, the electronic excess range Aeis_ex_i-1 causes the weight value Wois_i of the optical image stabilization weight Wois in the first imaging period T_img_i to be greater than the weight value Wois_i-1 in the second imaging period T_img_i-1. That is to say, the increase of the second magnitude can cause the weight value of the optical image stabilization to have an increasing change trend.
  • the processor component 90 may be further based on the angle ratio of the optical excess amplitude Aois_ex_i-1 in the field of view A_fov of the lens 10 where the optical mirror group 31 is located, and the electronic anti-shake weight Weis in the second imaging period T_img_i-1
  • the weight value Weis_i-1 determines the weight value Weis_i of the electronic anti-shake weight Weis in the first imaging period T_img_i, and, according to the angle of the electronic excess range Aeis_ex_i-1 in the field of view A_fov of the lens 10 where the optical mirror group 31 is located ratio, and the weight value Wois_i-1 of the OIS weight Wois in the second imaging period T_img_i ⁇ 1, determine the weight value Wois_i of the OIS weight Wois in the first imaging period T_img_i.
  • the determination method can be expressed as:
  • Wois_i Wois_i-1+Aeis_ex_i-1/A_fov;
  • Weis_i Weis_i-1+Aois_ex_i-1/A_fov.
  • the optical excess range Aois_ex_i-1 may include optical deflection excess range Yaw_ois_ex_i-1, optical pitch excess range Pitch_ois_ex_i-1, and optical flip excess range Roll_ois_ex_i-1
  • electronic excess range Aeis_ex_i-1 may specifically include electronic deflection excess range Yaw_eis_ex_i-1, electronic pitch excess range Pitch_eis_ex_i-1, and electronic rollover excess range Roll_eis_ex_i-1, and the above determination method may be specifically expressed as:
  • Wois_yaw_i Wois_yaw_i-1+Yaw_eis_ex_i-1/A_fov_yaw;
  • Weis_yaw_i Weis_yaw_i-1+Yaw_ois_ex_i-1/A_fov_yaw;
  • Wois_pitch_i Wois_pitch_i-1+Pitch_eis_ex_i-1/A_fov_pitch;
  • Weis_pitch_i Weis_pitch_i-1+Pitch_ois_ex_i-1/A_fov_pitch;
  • Wois_roll_i Wois_roll_i-1+Roll_eis_ex_i-1/A_fov_roll;
  • Weis_roll_i Weis_roll_i-1+Roll_ois_ex_i-1/A_fov_roll;
  • A_fov_yaw represents the horizontal field angle in the field of view A_fov
  • A_fov_pitch represents the vertical field of view in the field of view
  • A_fov_roll represents the angle range of the field of view A_fov (for example, 360°).
  • the anti-shake driving assembly 70 can realize the continuous angle correction of the imaging coordinate system by adjusting the relative pose between the optical mirror group 31 and the photosensitive element 32.
  • the optical anti-shake capability range can be It is determined according to the physical adjustment performance of the relative pose between the optical lens group 31 and the photosensitive element 32 .
  • the physical adjustment performance of the relative pose between the optical lens group 31 and the photosensitive element 32 may involve at least one of an adjustment rate of the relative pose and a pose limit. That is to say, the anti-shake driving assembly 70 can realize continuous angle correction of the real-time relative angle by adjusting the relative pose between the optical lens group 31 and the photosensitive element 32 .
  • the processor component 90 can be further used to determine the second single-frame equivalent spatial angle A_eq_i- 1
  • the required target speed Vobj_i-1 exceeds the speed difference (Vobj_i-1-Vmax) of the preset maximum speed Vmax for relative pose adjustment between the optical lens group 31 and the photosensitive element 32, and the optical excess range Aois_ex_i- 1 may include the magnitude that causes the speed difference in the second single-frame equivalent spatial angle A_eq_i-1, which may be referred to as the speed overrun magnitude Aois_v_ex_i-1.
  • the target velocity Vobj_i-1 indicates: in order to correct the imaging coordinate system from the current relative angle Aois_i-1_n to the target velocity ideal correction angle Aois_v_obj_i-1 of the opposite and equal amplitude to the second single frame equivalent space angle A_eq_i-1, The theoretical rate of change of the relative pose between the optical lens group 31 and the photosensitive element 32 within one frequency period T_gyro of the first frequency f_gyro.
  • the target velocity Vobj_i-1 ⁇ T_gyro ⁇ Kapm
  • Kapm is a preset conversion factor
  • the conversion factor Kapm is based on the relative relationship between the optical lens group 31 and the photosensitive element 32
  • the physical adjustment performance of the pose is determined, and represents the unit correction range of the imaging coordinate system that can be generated by adjusting the relative pose between the optical mirror group 31 and the photosensitive element 32 every preset unit step.
  • the preset maximum speed Vmax is determined according to the limit speed supported by the physical adjustment performance of the relative pose between the optical lens group 31 and the photosensitive element 32 .
  • the target speed Vobj_i-1 represents the target horizontal speed v_yaw_obj_i-1 corresponding to the yaw direction
  • the preset maximum speed Vmax represents the maximum horizontal speed v_yaw_max for the relative pose adjustment between the optical mirror group 31 and the photosensitive element 32
  • the target The ideal velocity correction angle Aois_v_obj_i-1 represents the target horizontal velocity ideal correction angle Yaw_ois_v_obj_i-1 of the opposite magnitude to the single-frame yaw equivalent spatial angle Yaw_eq_i-1 represented by the second single-frame equivalent spatial angle A_eq_i-1, then :
  • v_yaw_obj_i-1 ⁇ T_gyro ⁇ Kapm_yaw
  • the physical adjustment performance of the horizontal relative position is determined, and represents the unit deflection correction range of the imaging coordinate system that can be generated by adjusting the horizontal relative position between the optical mirror group 31 and the photosensitive element 32 per preset horizontal distance unit step.
  • the target speed Vobj_i-1 represents the target vertical speed v_pitch_obj_i-1 corresponding to the pitch direction
  • the preset maximum speed Vmax represents the maximum vertical speed v_pitch_max of the relative pose adjustment between the optical mirror group 31 and the photosensitive element 32
  • the target velocity ideal correction angle Aois_v_obj_i-1 represents the target vertical velocity ideal correction angle Pitch_ois_v_obj_i- 1, then:
  • v_pitch_obj_i-1 ⁇ T_gyro ⁇ Kapm_pitch
  • the target velocity Vobj_i-1 represents the target angular velocity ⁇ _roll_obj_i-1 corresponding to the flip direction
  • the preset maximum velocity Vmax represents the maximum angular velocity ⁇ _roll_max of the relative pose adjustment between the optical mirror group 31 and the photosensitive element 32
  • the target velocity is ideally corrected
  • the angle Aois_v_obj_i-1 represents the ideal flip correction angle Roll_ois_obj_i-1 of the target angular velocity that is reversed and equal to the single-frame flip equivalent space angle Roll_eq_i-1 represented by the second single-frame equivalent space angle A_eq_i-1, then:
  • ⁇ _roll_obj_i-1 ⁇ T_gyro ⁇ Kapm_roll
  • the physical adjustment performance of the relative angle is determined, and represents the unit inversion correction range of the imaging coordinate system that can be generated by adjusting the relative angle of translation between the optical mirror group 31 and the photosensitive element 32 by a preset angle unit step.
  • FIG. 6 is an exemplary effect diagram of the optical excess range determined based on the speed limit during the weight updating process shown in FIG. 5 .
  • Shown in Fig. 6 is the fitting curve W71 representing the ideal correction angle Aois_v_obj_i-1 of the target speed, and the fitting curve W72 representing the speed overrun range Aois_v_ex_i-1, as can be seen from Fig. 6:
  • the imaging coordinate system is corrected from the current relative angle Aois_i-1_n to the ideal correction angle Aois_v_obj_i-1 at the target speed Vobj_i-1 within a frequency cycle T_gyro to the ideal correction angle Aois_v_obj_i-1 of the target speed.
  • Aois_v_ex_i-1
  • the target horizontal velocity ideal correction stroke ⁇ Yaw_ois_v_obj_i-1
  • , and the limit horizontal speed correction stroke ⁇ Yaw_ois_v_max supported by the maximum horizontal speed v_yaw_max is determined as v_yaw_max ⁇ T_gyro ⁇ Kapm_yaw:
  • the overrun range Yaw_ois_v_ex_i-1 may include the difference between the target horizontal speed ideal correction stroke ⁇ Yaw_ois_v_obj_i-1 and the limit horizontal speed correction stroke ⁇ Yaw_ois_v_max, otherwise, the yaw speed overrun range Yaw_ois_v_ex_i-1 is 0.
  • the target vertical velocity ideal correction stroke ⁇ Pitch_ois_v_obj_i-1 from the current pitch relative angle Pitch_ois_i-1_n to the target vertical velocity ideal correction angle Pitch_ois_v_obj_i-1
  • the vertical speed exceeding range Pitch_ois_v_ex_i-1 may include the difference between the target vertical speed ideal correction stroke ⁇ Pitch_ois_v_obj_i-1 and the limit vertical speed correction stroke ⁇ Pitch_ois_v_max, otherwise, the vertical speed exceeding range Pitch_ois_v_ex_i-1 is 0.
  • the target angular velocity ideal correction stroke ⁇ Roll_ois_ ⁇ _obj_i-1
  • Roll_ois_ ⁇ _obj_i-1-Roll_ois_i is corrected from the current flip relative angle Roll_ois_i-1_n to the target angular velocity ideal correction angle Roll_ois_ ⁇ _obj_i-1 -1_n
  • the limit angular velocity correction stroke ⁇ Roll_ois_ ⁇ _max supported by the maximum angular velocity ⁇ _roll_max is determined as ⁇ _roll_max ⁇ T_gyro ⁇ Kapm_roll):
  • the limit range Roll_ois_ ⁇ _ex_i-1 may include the difference between the target angular velocity ideal correction stroke ⁇ Roll_ois_ ⁇ _obj_i-1 and the limit angular velocity correction stroke ⁇ Roll_ois_ ⁇ _max, otherwise, the angular velocity excess range Roll_ois_ ⁇ _ex_i-1 is 0.
  • the processor component 90 can be further used to determine the second single-frame equivalent spatial angle A_eq_i -1
  • the required target pose Pobj_i-1 exceeds the pose difference outside the preset pose range Pmax (ie Pobj_i-1-Pmax), and the optical excess range Aois_ex_i-1 can include the second single frame equivalent space
  • the magnitude of the pose difference in the angle A_eq_i-1 may be called the pose excess magnitude Aois_p_ex_i-1.
  • the target pose Pobj_i-1 indicates that in order to make the imaging coordinate system reach the ideal correction angle Aois_p_obj_i-1 of the target pose opposite to the second single frame equivalent space angle A_eq_i-1, the optical mirror group 31 and the photosensitive The theoretical relative pose after the relative pose between the elements 32 has been adjusted.
  • the preset pose range Pmax is determined according to the limit pose that can be supported by the physical adjustment performance of the relative pose between the optical mirror group 31 and the photosensitive element 32.
  • the preset pose range Pmax can be smaller than the physical adjustment The physical adjustment range of the limit pose that the performance can support, so as to avoid the relative pose between the optical lens group 31 and the photosensitive element 32 being at the boundary of the physical adjustment range (the imaging quality during the boundary is relatively low).
  • the target pose Pobj_i-1 can represent the target horizontal position p_yaw_obj_i-1 corresponding to the yaw direction, or the target vertical position p_pitch_obj_i-1, or the target roll angle p_roll_obj_i-1, correspondingly, the target pose ideal correction angle Aois_p_obj_i-1 can represent:
  • the target horizontal position ideal correction angle Yaw_ois_p_obj_i-1 is opposite to the single-frame yaw equivalent spatial angle Yaw_eq_i-1 represented by the second single-frame equivalent spatial angle A_eq_i-1; or,
  • FIG. 7 is an exemplary effect diagram of the optical excess range determined based on the pose limit during the weight update process as shown in FIG. 5 .
  • the fitting broken line W80 shown in FIG. 7 represents the target pose correction range Aois_p_ex_i-1
  • the target pose correction range Aois_p_ex_i-1 is the target pose ideal correction angle Aois_p_obj_i-1 (for example, 2°) exceeds the preset Part of the limit correction angle Aois_p_max (for example, 1.8°) that the pose range Pmax can support.
  • Aois_p_ex_i-1 Aois_p_obj_i-1-Aois_p_max;
  • the target horizontal position p_yaw_obj_i-1 exceeds the limit horizontal position p_yaw_max in the preset pose range Pmax, resulting in the ideal correction angle Yaw_ois_p_obj_i-1 of the target horizontal position in the imaging coordinate system, which exceeds the limit that the preset pose range Pmax can support
  • the horizontal position overrun Yaw_ois_p_ex_i-1 is set to 0.
  • the target vertical position p_pitch_obj_i-1 exceeds the limit vertical position p_pitch_max in the preset pose range Pmax, resulting in the ideal correction angle Pitch_ois_p_obj_i-1 of the target vertical position in the imaging coordinate system, which exceeds the preset pose range Pmax.
  • optical excess range Aois_ex_i-1 Aois_v_ex_i-1 determined based on the speed limit as shown in Figure 6
  • Weis_i Weis_i-1+(Aois_v_ex_i-1+Aois_p_ex_i-1)/A_fov, namely:
  • Weis_yaw_i Weis_yaw_i-1+(Yaw_ois_v_ex_i-1+Yaw_ois_p_ex_i-1)/A_fov_yaw;
  • Weis_pitch_i Weis_pitch_i-1+(Pitch_ois_v_ex_i-1+Pitch_ois_p_ex_i-1)/A_fov_pitch;
  • Weis_roll_i Weis_roll_i-1+(Roll_ois_v_ex_i-1+Roll_ois_p_ex_i-1)/A_fov_roll.
  • the processor component 90 is used to determine the inter-frame deviation of the electronic image stabilization compensation angle of the second imaging period T_img_i-1 compared to the previous third imaging period T_img_i-2 ⁇ Aeis_i-1 (i.e.
  • the second imaging period T_img_i-1's electronic image stabilization pitch compensation angle inter-frame deviation ⁇ Pitch_eis_i-1 ie
  • Pitch_eis_ex_i-1 ⁇ Pitch_eis_i-1- ⁇ Pitch_int_th of pitch jitter threshold Pitch_int_th;
  • Roll_eis_ex_i-1 ⁇ Roll_eis_i-1- ⁇ Roll_int_th of rollover jitter exceeding limit of threshold Roll_int_th.
  • the processor component 90 can be further used to obtain the angle threshold value associated with the shutter time t_sh of the lens 10 where the optical lens group 31 is located, which can be called the anti-blur angle threshold value A_ed(sh), which is used to determine the electronic excess
  • the inter-frame jitter threshold A_int_th of the amplitude Aeis_ex_i-1, that is, the inter-frame jitter threshold A_int_th may include the obtained anti-blur angle threshold A_ed(sh), and the anti-blur angle threshold A_ed(sh) is based on the Time t_sh is determined by the associated sharpness threshold.
  • the sharpness of the image 30 may be lower than the preset sharpness threshold
  • the electronic excess amplitude Aeis_ex_i-1 ⁇ Aeis_i-1-A_ed(sh)
  • the weight value Wois_i of the optical image stabilization weight Wois in the first imaging period T_img_i can be updated as:
  • Wois_i Wois_i-1+( ⁇ Aeis_i-1-A_ed(sh))/A_fov;
  • the anti-blur angle threshold A_ed(sh) may represent the anti-blur yaw angle threshold Yaw_ed(sh), or the anti-blur pitch angle threshold Pitch_ed(sh), or the anti-blur roll angle threshold Roll_ed(sh) in the yaw direction.
  • FIG. 8 is an exemplary effect diagram of the electronic excess range determined in relation to the shutter time during the weight update process as shown in FIG. 5 .
  • broken line W91 represents the second electronic image stabilization compensation angle Aeis_i-1 of the second imaging period T_img_i-1 and the third electronic image stabilization compensation angle Aeis_i-2 of the third imaging period T_img_i-2
  • broken line W92 represents the electronic image stabilization compensation angle Aeis_i-1 of the third imaging period T_img_i-2.
  • the inter-frame deviation of the anti-shake compensation angle ⁇ Aeis_i-1, and the broken line W90 indicates the electronic excess range Aeis_ex_i-1 of the electronic anti-shake compensation angle inter-frame deviation ⁇ Aeis_i-1 exceeding the anti-blur angle threshold A_ed(sh).
  • the processor component 90 can be further used to determine the weight values Weis_i+1 and The weight value Weis_i+1 of the electronic image stabilization weight Weis in the fourth imaging period T_img_i+1 will not be described in detail herein.
  • Fig. 9 is a schematic flowchart of an exemplary anti-shake control method in another embodiment of the present application.
  • the control method provided in another embodiment may be periodically executed by the processor component of the camera at a second frequency (for example, the periodic frequency f_img at which the photosensitive element generates an image), and the control method may include :
  • S900 Determine the weight value of the optical anti-shake weight and the electronic anti-shake weight in the first imaging period of the photosensitive element for realizing cooperative anti-shake, and the weight of the optical anti-shake weight in the first imaging period of the photosensitive element of the camera
  • the value is intended to limit the real-time relative angle of the photosensitive element in the camera body coordinate system to the scope of the camera's optical image stabilization capability, so that the real-time relative angle in the first imaging cycle has a relative angle relative to the camera in the space coordinate system The real-time angle difference of the real-time space angle of .
  • the real-time spatial angle of the camera in the spatial coordinate system may be determined according to the real-time shaking data continuously generated by a shaking sensor such as a gyroscope at a first frequency f_gyro, and the first frequency f_gyro of the shaking sensor generating real-time shaking data is, Not less than twice the second frequency f_img of the photosensitive element imaging.
  • a shaking sensor such as a gyroscope at a first frequency f_gyro
  • the first frequency f_gyro of the shaking sensor generating real-time shaking data is, Not less than twice the second frequency f_img of the photosensitive element imaging.
  • S910 Determine a first single-frame equivalent spatial angle in the first imaging period according to the real-time spatial angle in the first imaging period.
  • the first single-frame equivalent spatial angle may be a peak angle among real-time spatial angles within the first imaging period.
  • S930 Determine the first electronic anti-shake compensation angle according to the weight value of the electronic anti-shake weight in the first imaging period and the first single-frame equivalent space angle, so as to use the determined first electronic anti-shake compensation angle
  • Image processing is performed on the image obtained at the first refresh moment at the end of the first imaging period, and electronic anti-shake is realized by making the image generate pixel array adjustment for compensating real-time angle difference.
  • next imaging cycle can be used as the first imaging cycle, and return to S900 to continue the above process.
  • the real-time spatial angle of the camera can only be offset by the optical image stabilization to a part within the range of the optical image stabilization capability of the camera, so as to help reduce the risk of optical image stabilization failure; and , based on a part of the spatial angle offset by optical image stabilization, it can also reduce the risk of image clarity being too low due to too large a spatial angle.
  • Image processing is performed on the improved image after optical image stabilization, which can compensate the impact of the remaining part of the real-time spatial angle of the camera on the image quality. Furthermore, through the dual-mode combination of electronic image stabilization and optical image stabilization, it helps to reduce the risk of image stabilization failure.
  • S900 in the above process may occur before the first imaging period, and may specifically include:
  • the real-time space angle and the second electronic anti-shake compensation angle in the second imaging period before the first imaging period determine the weight value of the optical anti-shake weight and the electronic anti-shake weight in the first imaging period, so as to promote:
  • the real-time relative angle is intended to be limited to converge within a preset angle correction limit range, which may be determined according to the optical image stabilization capability range;
  • the sharpness of the image obtained at the first refreshing moment is not lower than the preset sharpness threshold, which may be determined according to the preset minimum image sharpness for avoiding electronic anti-shake failure.
  • the real-time spatial angle in the second imaging period determine the second single-frame equivalent spatial angle of the second imaging period, and determine the optical excess range of the second single-frame equivalent spatial angle beyond the angle correction limit range, wherein the optical The excess range produces a weight change trend that increases the weight value of the electronic anti-shake weight;
  • optical image stabilization continuously corrects the real-time relative angle of the imaging coordinate system, it is realized by adjusting the relative pose between the optical lens group and the photosensitive element of the camera, and the range of optical image stabilization capability is based on the optical lens group and photosensitive element.
  • the physical adjustment performance of the relative pose between the elements is determined, then, when determining the optical excess range of the second single frame equivalent space angle exceeding the angle correction limit range:
  • the optical excess range may include the speed limit range that causes the speed difference in the second single-frame equivalent space angle
  • the target velocity represents the theory of the relative pose between the optical mirror group and the photosensitive element in order to correct the imaging coordinate system from the current relative angle to the ideal correction angle of the target velocity that is opposite to the equivalent spatial angle of the second single frame and equal in magnitude rate of change (for example, a theoretical rate of change within one frequency cycle of the first frequency); and, the preset maximum speed may be determined according to the limit speed supported by the physical adjustment performance; and/or,
  • the target pose required by the second single-frame equivalent space angle exceeds the pose difference outside the preset pose range, wherein the optical excess range includes the second single-frame equivalent space angle that causes the pose difference Pose overrun range;
  • the target pose represents the theoretical relative pose after the relative pose has been adjusted in order to make the imaging coordinate system reach the ideal correction angle of the target pose with the opposite and equal amplitude to the equivalent space angle of the second single frame ;
  • the range of preset poses can be determined according to the limit poses supported by the physical adjustment performance.
  • control method may further include: obtaining an anti-blur angle threshold that changes in association with the shutter time of the lens where the optical mirror group is located, wherein the inter-frame shake amplitude threshold used to determine the electronic excess amplitude may include the obtained anti-blur angle threshold A blur angle threshold, and the anti-blur angle threshold may be determined according to the sharpness threshold associated with the shutter time.
  • a control device including: a processor; a memory for storing instructions executable by the processor; wherein the processor is configured to execute the steps of any one of the control methods described above.
  • a computer-readable storage medium where instructions are stored on the computer-readable storage medium, and when the instructions are executed by a processor, the steps of any one of the control methods described above are implemented.
  • a computer program product including instructions, which, when running on a processor, causes the processor to execute the steps of any one of the control methods described above.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Studio Devices (AREA)
  • Adjustment Of Camera Lenses (AREA)

Abstract

一种用于利用光学防抖和电子防抖实现协同防抖的摄像机。基于本申请，在感光元件的任意一个刷新周内，摄像机的实时空间角度可以只有在摄像机的光学防抖能力范围内的一部分被光学防抖抵消，以助于降低光学防抖失效的风险；并且，基于光学防抖所抵消的一部分空间角度，还可以降低图像清晰度由于空间角度过大而过低的风险，从而，在该成像周期结束时的刷新时刻，通过电子防抖对具有清晰度被光学防抖改善后的图像实施图像处理，可以补偿摄像机的实时空间角度的剩余部分对图像质量的影响。进而，通过以电子防抖补偿光学防抖的双模式结合，有助于降低防抖失效的风险。

Description

摄像机
本申请要求于2021年07月21日提交中国专利局、申请号为202110822687.7发明名称为“用于利用光学防抖和电子防抖实现协同防抖的摄像机”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。
技术领域
本申请涉及摄像机的防抖技术，特别涉及一种摄像机。
背景技术
通常情况下，由于光学防抖和电子防抖的原理不同，因而这两种防抖模式是择一使用的。但光学防抖和电子防抖均存在各自的性能瓶颈，因此，无论使用哪一种防抖模式，都有可能发生防抖失效。
可见，如何降低防抖失效的风险，成为现有技术中有待解决的技术问题。
发明内容
有鉴于此，本申请的实施例提供了一种摄像机，有助于降低防抖失效的风险。
在一个实施例中，该摄像机可以包括：
抖动传感器，用于以第一频率持续产生所述摄像机的实时抖动数据，其中，所述第一频率不低于第二频率的两倍，所述第二频率为所述摄像机的感光元件成像的频率，并且，所述实时抖动数据用于确定所述摄像机在空间坐标系中的实时空间角度；
处理器组件，用于以所述第二频率更新光学防抖权重值和电子防抖权重值；
防抖驱动组件，用于根据所述光学防抖权重值，对所述感光元件的成像坐标系在所述摄像机的机身坐标系中的实时相对角度，实施与所述实时空间角度反向的持续地角度校正，以实现响应于所述实时空间角度的光学防抖；
其中，根据所述感光元件的第一成像周期内的光学防抖权重值对实时相对角度进行角度校正，使得角度校正后的实时相对角度在所述摄像机的光学防抖能力范围内，以使得角度校正后的实时相对角度具有相对于实时空间角度的实时角度差值；
并且，处理器组件还用于根据所述第一成像周期内的实时空间角度，确定所述第一成像周期的等效空间角度，作为第一等效空间角度，并且，根据所述第一成像周期内的电子防抖权重值、以及所述第一等效空间角度，确定第一电子防抖补偿角度，以根据所述第一电子防抖补偿角度调节第一刷新时刻得到的图像的像素阵列，得到像素阵列的调节幅度，所述像素阵列的调节幅度用于补偿所述实时角度差值，实现电子防抖；所述第一刷新时刻表示所述第一成像周期的结束时刻。
在另一个实施例中，提供了一种防抖的控制方法，包括：
确定摄像机的感光元件的第一成像周期内的光学防抖权重值和电子防抖权重值，其中，所述第一成像周期内的光学防抖权重值，用于对所述感光元件的成像坐标系在所述摄像机的机身坐标系中的实时相对角度，实施与所述摄像机在空间坐标系中的实时空间角度反向的持续地角度校正，使得角度校正后的实时相对角度在所述摄像机的光学防抖能力范围内，以使得角度校正后的实时相对角度具有相对于实时空间角度的实时角度差值；
根据所述第一成像周期内的实时空间角度，确定所述第一成像周期的等效空间角度，作为第一等效空间角度；以及，
根据所述第一成像周期内的电子防抖权重值、以及所述第一等效空间角度，确定第一电子防抖补偿角度，以根据所述第一电子防抖补偿角度调节第一刷新时刻得到的图像的像素阵列，得到像素 阵列的调节幅度，所述像素阵列的调节幅度用于补偿所述实时角度差值，实现电子防抖。
在另一个实施例中，提供了一种控制设备，包括：处理器；用于存储处理器可执行指令的存储器；其中，所述处理器被配置为执行上述任一所述的控制方法的步骤。
在另一个实施例中，提供了一种计算机可读存储介质，所述计算机可读存储介质上存储有指令，所述指令被处理器执行时实现上述任一所述的控制方法的步骤。
在另一个实施例中，提供了一种包含指令的计算机程序产品，当其在处理器上运行时，使得处理器执行上述任一所述的控制方法的步骤。
基于上述实施例，在感光元件的任意一个刷新周期内，摄像机的实时空间角度可以只有在摄像机的光学防抖能力范围内的一部分被光学防抖抵消，以助于降低光学防抖失效的风险；并且，基于光学防抖所抵消的一部分空间角度，还可以降低图像清晰度由于空间角度过大而过低的风险，从而，在该成像周期结束时的刷新时刻，通过电子防抖对具有清晰度被光学防抖改善后的图像实施图像处理，可以补偿摄像机的实时空间角度的剩余部分对图像质量的影响。进而，通过以电子防抖补偿光学防抖的双模式结合，有助于降低防抖失效的风险。
附图说明
为了更清楚地说明本申请实施例和现有技术的技术方案，下面对实施例和现有技术中所需要使用的附图作简单地介绍，显而易见地，下面描述中的附图仅仅是本申请的一些实施例，对于本领域普通技术人员来讲，在不付出创造性劳动的前提下，还可以根据这些附图获得其他的附图。
图1为本申请的一个实施例中的摄像机的硬件架构示意图；
图2为如图1所示摄像机确定的实时空间角度的示例性示意图；
图3为如图1所示摄像机实现的协同防抖的原理示意图；
图4为如图3所示协同防抖采用的电子防抖补偿光学防抖的示例性效果图；
图5为适用于如图3所示协同防抖的权重更新过程的原理示意图；
图6为如图5所示权重更新过程中基于速度极限确定的光学超额幅度的示例性效果图；
图7为如图5所示权重更新过程中基于位姿极限确定的光学超额幅度的示例性效果图；
图8为如图5所示权重更新过程中关联快门时间确定的电子超额幅度的示例性效果图；
图9为本申请的另一个实施例中防抖的控制方法的示例性流程示意图。
具体实施方式
为使本申请的目的、技术方案、及优点更加清楚明白，以下参照附图并举实施例，对本申请进一步详细说明。显然，所描述的实施例仅仅是本申请一部分实施例，而不是全部的实施例。基于本申请中的实施例，本领域普通技术人员在没有作出创造性劳动前提下所获得的所有其他实施例，都属于本申请保护的范围。
图1为本申请的一个实施例中的摄像机的硬件架构示意图。请参见图1，该摄像机可以包括镜头10和机身20，其中，镜头10具有光学镜组31，机身20中具有感光元件32、以及防抖驱动组件70和处理器组件90，并且，镜头10或机身20中还可以装设有抖动传感器50。
感光元件32可以包括CMOS(Complementary Metal Oxide Semiconductor，互补金属氧化物半导体)或CCD(Charge Coupled Device，电荷耦合器件)，用于在光学镜组31的成像路径中曝光成像，并刷新输出曝光成像得到的图像。
抖动传感器50用于以第一频率f_gyro持续产生摄像机的实时抖动数据，例如，抖动传感器50可以包括例如陀螺仪等用于确定摄像机的实时空间角度的传感器，即，该实时空间角度可以认为是摄像机的机身坐标系在空间坐标系中的实时角度。
其中，抖动传感器50以第一频率f_gyro持续产生的实时抖动数据用于确定摄像机在空间坐标系中的实时空间角度。并且，抖动传感器50产生实时抖动数据的第一频率f_gyro，不低于感光元件32成像的第二频率f_img的两倍。因此，在感光元件32的每个成像周期T_img_i内，实时空间角度包括根据至少两次实时抖动数据确定的角度值序列{A_i_1，…，A_i_n}，i表示感光元件32的第i个成像周期T_img_i的序列号，i为大于0的正整数，并且，n为一个成像周期T_img_i内的实时抖动数据的产生次数，n大于等于2。第二频率f_img也就是摄像机生成的视频图像的帧率。
摄像机的抖动包括水平摆动方向上的偏摆抖动、仰俯摆动方向上的仰俯抖动、以及绕镜头光轴的旋转方向上的翻转抖动。
相应地，抖动传感器50产生的实时抖动数据可以是响应于偏摆抖动、仰俯抖动以及翻转抖动中的至少之一而被引发的。以抖动传感器50包括陀螺仪(诸如3轴陀螺仪)为例，抖动传感器50以第一频率f_gyro持续产生的实时抖动数据可以包括响应于摄像机的偏摆抖动、仰俯抖动、以及翻转抖动中的至少之一的角速度或角加速度，并且，通过预设的算法，可以根据角速度或角加速度确定摄像机在空间坐标系中的实时空间角度{A_i_1，…，A_i_n}。
从而，摄像机的实时空间角度可以包括反应偏摆抖动的偏摆空间角度、反应仰俯抖动的仰俯空间角度、以及反应翻转抖动的翻转空间角度。即，前文用于表示一个成像周期T_img_i内的实时空间角度所使用的通用表达式{A_i_1，…，A_i_n}，可以具体化为：实时偏摆空间角度{Yaw_i_1，…，Yaw_i_n}，或实时仰俯空间角度{Pitch_i_1，…，Pitch_i_n}，或实时翻转空间角度{Roll_i_1，…，Roll_i_n}。
图2为如图1所示摄像机确定的实时空间角度的示例性示意图。在图2中示出了偏摆抖动、仰俯抖动以及翻转抖动中的任一个发生时，在对应的抖动方向上的实时空间角度{A_i_1，…，A_i_n}的变化状态，实时空间角度{A_i_1，…，A_i_n}在图2中以第一频率f_gyro的频率周期Tgyro为间隔离散分布，并且，图2中还示出了离散分布的实时空间角度{A_i_1，…，A_i_n}的拟合曲线W20，以表示实时空间角度{A_i_1，…，A_i_n}的变化趋势。
在该实施例中，根据抖动传感器50产生的实时抖动数据确定的实时空间角度{A_i_1，…，A_i_n}，是相对于预先标定的空间基准角度A_abs_ref的角度值，该空间基准角度A_abs_ref是在空间坐标系中标定的，例如，该空间基准角度A_abs_ref可以表示镜头光轴在摄像机的部署空间中的期望角度。即，实时偏摆空间角度{Yaw_i_1，…，Yaw_i_n}均是相对于预先标定的空间偏摆基准角度Yaw_abs_ref的角度值，实时仰俯空间角度{Pitch_i_1，…，Pitch_i_n}均是相对于预先标定的空间仰俯基准角度Pitch_abs_ref的角度值，以及实时翻转空间角度{Roll_i_1，…，Roll_i_n}均是相对于预先标定的空间翻转基准角度Roll_abs_ref的角度值。
在该实施例中，每个成像周期T_img_i内，可以至少针对偏摆抖动和仰俯抖动实施光学防抖和电子防抖的协同防抖；而翻转抖动则可以完全依靠电子防抖，或者，也可以针对翻转抖动实施光学防抖和电子防抖的协同防抖。
防抖驱动组件70可以用于实现光学防抖，具体地，防抖驱动组件70可以用于对感光元件32的成像坐标系在摄像机的机身坐标系中的实时相对角度，实施与实时空间角度{A_i_1，…，A_i_n}反向的持续角度校正，即，防抖驱动组件70可以实施持续地角度校正，且该角度校正的方向与实时空间角度反向，以实现响应于实时空间角度{A_i_1，…，A_i_n}的光学防抖。其中，感光元件32的成像坐标系的实时相对角度，可以由光学镜组31与感光元件32之间的相对位姿确定，因此，防抖驱动组件70可以通过调节摄像机的光学镜组31与感光元件32之间的相对位姿，实现对成像坐标系的持续角度校正。例如，防抖驱动组件70可以包括用于驱动光学镜组31与感光元件32之间的相对位姿调节的OIS驱动器。
具体地，防抖驱动组件70实施的持续角度校正，可以是响应于偏摆抖动、仰俯抖动以及翻转抖动中的至少之一而被引发的。
其中，针对偏摆抖动和仰俯抖动，防抖驱动组件70调节的相对位姿可以包括对光学镜组31与感光元件32之间的相对平移位置的调节。
例如，在摄像机发生偏摆抖动或仰俯抖动的期间内，通过驱动光学镜组31相对于感光元件32的水平平移或竖直平移，防抖驱动组件70可以调节光学镜组31与感光元件32之间的相对水平位置或相对竖直位置，从而使成像坐标系在摄像机的偏摆抖动或仰俯抖动相反的方向上，摆动至相对于摄像机的指定水平相对角度或指定竖直相对角度，以缓解成像坐标系跟随于摄像机的偏摆抖动或仰俯抖动(即机身坐标系在空间坐标系中的偏摆抖动或仰俯抖动)的波动，并促使成像坐标系在空间坐标系中的空间角度保持稳定。通过光学镜组31的平移实现的光学防抖，也可以称为镜头防抖或光学镜组位移防抖。
再例如，作为镜头防抖或光学镜组位移防抖的替代方案，通过驱动感光元件32相对于光学镜组31的水平平移或竖直平移，防抖驱动组件70同样可以调节光学镜组31与感光元件32之间的相对水平位置或相对竖直位置，并以此缓解成像坐标系跟随于摄像机的偏摆抖动或仰俯抖动的波动。通过感光元件32的移动实现的光学防抖，也可以称为机身防抖或感光元件位移防抖。
并且，若针对翻转抖动也实施光学防抖和电子防抖的协同防抖，则，防抖驱动组件70调节的相对位姿可以包括对光学镜组31与感光元件32之间的相对平转角度的调节。
例如，感光元件32可转动地装设在摄像机的机身20内，通过驱动感光元件32相对于光学镜组31的平转，防抖驱动组件70可以调节光学镜组31与感光元件32之间的相对平转角度，从而使成像坐标系在摄像机的翻转抖动相反的方向上，翻转至相对于摄像机的指定翻转相对角度，以缓解成像坐标系跟随于摄像机的翻转抖动(即机身坐标系在空间坐标系中的翻转抖动)的波动，并促使成像坐标系在空间坐标系中的空间角度保持稳定。
处理器组件90可以用于实现电子防抖。具体地，处理器组件90可以用于对感光元件32在任意成像周期T_img_i结束时的刷新时刻输出的图像进行图像处理，以通过图像处理产生的像素阵列调节实现电子防抖。例如，处理器组件90可以包括用于对图像进行图像处理的图像处理器。也就是说，处理器组件90可以对感光元件32在任意成像周期T_img_i结束时的刷新时刻输出的图像的像素阵列进行调节，得到像素阵列的调节幅度，基于该调节幅度能够实现电子防抖。
其中，处理器组件90通过图像处理对像素阵列的调节，可以是响应于偏摆抖动、仰俯抖动以及翻转抖动中的至少之一而被引发的。即，对于偏摆抖动和仰俯抖动的电子防抖，可以通过处理器组件90对感光元件32输出的图像进行像素阵列平移；对于翻转抖动的电子防抖，处理器组件90可以对感光元件32输出的图像进行像素阵列平转。
不同于光学防抖直接利用持续的实时空间角度{A_i_1，…，A_i_n}对实时相对角度持续实施角度校正的方式，电子防抖是在一个成像周期T_img_i结束时的刷新时刻对图像的像素阵列实施单次处理，因此，在任意成像周期T_img_i的刷新时刻执行的电子防抖，可以使用与该成像周期T_img_i内的实时空间角度{A_i_1，…，A_i_n}关联的单帧等效空间角度A_eq_i，即，与实时偏摆空间角度{Yaw_i_1，…，Yaw_i_n}关联的单帧等效偏摆空间角度Yaw_eq_i，与实时仰俯空间角度{Pitch_i_1，…，Pitch_i_n}关联的单帧等效仰俯空间角度Pitch_eq_i，以及，与实时翻转空间角度{Roll_i_1，…，Roll_i_n}关联的单帧等效翻转空间角度Roll_eq_i。单帧等效空间角度A_eq_i是与该成像周期T_img_i内的实时空间角度{A_i_1，…，A_i_n}相关联的，而该成像周期T_img_i表示一个图像帧的时长，即，单帧等效空间角度A_eq_i是与该图像帧对应的，因此可以称为单帧等效空间角度，在本申请中也可以称为等效空间角度。
具体地，处理器组件90可以从任意一个成像周期T_img_i内的实时空间角度{A_i_1，…，A_i_n}中采样得到该成像周期T_img_i的单帧等效空间角度A_eq_i，即，单帧等效空间角度A_eq_i也可以称为采样空间角度。
例如，处理器组件90可以从任意一个成像周期T_img_i内的实时空间角度{A_i_1，…，A_i_n}中采样得到峰值角度A_i_max(其角度绝对值最大)、或均值角度A_i_avg(角度绝对值的平均值)、或中位角度A_i_med(角度绝对值的中位值)，并且，将该成像周期T_img_i内的实时空间角度{A_i_1，…，A_i_n}中的峰值角度A_i_max、或均值角度A_i_avg、或中位角度A_i_med确定为该成像周期T_img_i内的单帧等效空间角度A_eq_i，即：
在单帧等效空间角度A_eq_i表示单帧等效偏摆空间角度Yaw_eq_i的情况下，实时偏摆空间角度{Yaw_i_1，…，Yaw_i_n}中的偏摆峰值角度Yaw_i_max、或偏摆均值角度Yaw_i_avg、或偏摆中位角度Yaw_i_med可以被确定为单帧等效偏摆空间角度Yaw_eq_i；
在单帧等效空间角度A_eq_i表示单帧等效仰俯空间角度Pitch_eq_i的情况下，实时仰俯空间角度{Pitch_i_1，…，Pitch_i_n}中的仰俯峰值角度Pitch_i_max、或仰俯均值角度Pitch_i_avg、或仰俯中位角度Pitch_i_med可以被确定为单帧等效仰俯空间角度Pitch_eq_i；
在单帧等效空间角度A_eq_i表示单帧等效翻转空间角度Roll_eq_i的情况下，实时翻转空间角度{Roll_i_1，…，Roll_i_n}中的翻转峰值角度Roll_i_max、或翻转均值角度Roll_i_avg、或翻转中位角度Roll_i_med可以被确定为单帧等效翻转空间角度Roll_eq_i。
由于峰值角度A_i_max能够反应出该成像周期T_img_i内曾发生过的最大抖动幅度，因此，优选地，可以将峰值角度A_i_max确定为单帧等效空间角度A_eq_i，以助于电子防抖使图像的像素阵列产生的调节幅度足以减轻甚至消除最大抖动幅度对图像质量的影响。
处理器组件90还用于实现光学防抖与电子防抖之间的协同控制。例如，该协同控制可以由处理器组件90中的图像处理器承担、或者也可以由处理器组件90中进一步包括的独立于图像处理器的另一处理器承担。
对于光学防抖和电子防抖的协同控制，该实施例中为摄像机提供了用于实现协同防抖的两个参数，即，光学防抖权重Wois和电子防抖权重Weis。其中，光学防抖权重Wois的权重值Wois_i以及电子防抖权重Weis的权重值Weis_i是可以与感光元件32的图像刷新同步更新调节的。光学防抖权重Wois的权重值Wois_i可以称为光学防抖权重值，电子防抖权重Weis的权重值Weis_i可以称为电子防抖权重值。
处理器组件90可以用于以第二频率f_img更新用于实现协同防抖的光学防抖权重Wois的权重值Wois_i、以及电子防抖权重Weis的权重值Weis_i，更新的依据可以是单个成像周期T_img_i内的实时空间角度{A_i_1，…，A_i_n}的单帧等效空间角度A_eq_i。
即，处理器组件90对光学防抖权重Wois的权重值Wois_i、以及电子防抖权重Weis的权重值Weis_i的更新，可以是响应于偏摆抖动、仰俯抖动以及翻转抖动中的至少之一而被引发的，在此情况下：
光学防抖权重Wois和电子防抖权重Weis可以表示对应实时偏摆空间角度{Yaw_i_1，…，Yaw_i_n}的光学偏摆防抖权重Wois_yaw和电子偏摆防抖权重Weis_yaw；
光学防抖权重Wois和电子防抖权重Weis还可以表示对应实时仰俯空间角度{Pitch_i_1，…，Pitch_i_n}的光学仰俯防抖权重Wois_pitch和电子仰俯防抖权重Weis_pitch；
若针对实时翻转空间角度{Roll_i_1，…，Roll_i_n}也实施光学防抖和电子防抖的协同防抖，则，光学防抖权重Wois和电子防抖权重Weis还可以表示对应实时翻转空间角度{Roll_i_1，…，Roll_i_n}的光学翻转防抖权重Wois_roll和电子翻转防抖权重Weis_roll。
处理器组件90对光学偏摆防抖权重Wois_yaw和电子偏摆防抖权重Weis_yaw的权重值更新，对光学仰俯防抖权重Wois_pitch和电子仰俯防抖权重Weis_pitch的权重更新，以及，对光学翻转防抖权重Wois_roll和电子翻转防抖权重Weis_roll的权重值更新，可以是相互独立的。
图3为如图1所示摄像机实现的协同防抖的原理示意图。请参见图3：
根据抖动传感器50产生的实时抖动数据确定实时空间角度{A_i_1，…，A_i_n}。防抖驱动组件70可以具体用于在任意一个成像周期T_img_i内，根据光学防抖权重Wois在该成像周期T_img_i内的权重值Wois_i，对感光元件32的成像坐标系相对于摄像机的实时相对角度，实施与实时空间角度{A_i_1，…，A_i_n}反向的持续角度校正，以实现响应于实时空间角度{A_i_1，…，A_i_n}的光学防抖。
也就是，在任意一个成像周期T_img_i内，成像坐标系相对于摄像机(即机身坐标系)的实时相对角度被校正为{Aosi_i_1，…，Aois_i_n}。
其中，Aois_i_j＝A_i_j×Wois_i/(Wois_i+Weis_i)，j为一个成像周期T_img_i内的一次实时抖动数据或实时空间角度的序列中基于时间的排序位置，j大于等于1且小于等于前文提及的n。
被校正后的实时相对角度Aois_i_j是指，为抵消摄像机抖动时所处的实时空间角度{A_i_1，…，A_i_n}所引发的成像坐标系相对于空间基准角度A_abs_ref(即空间坐标系中的空间基准角度)的偏差，而使成像坐标系相对于预先标定的相对基准角度A_rlv_ref的角度值，该相对基准角度A_rlv_ref可以是在摄像机的机身坐标系中标定的，例如，该相对基准角度A_rlv_ref可以表示镜头光轴在机身坐标系中的装配角度，并且，该相对基准角度A_rlv_ref在空间坐标系中的方位角度是响应于实时空间角度{A_i_1，…，A_i_n}变化的。
在任意一个成像周期T_img_i结束时的刷新时刻，电子防抖对图像实施的像素阵列调节所依据的电子防抖补偿角度Aeis_i＝A_eq_i×Weis_i/(Wois_i+Weis_i)。
相应地，处理器组件90可以具体用于根据任意一个成像周期T_img_i内的实时空间角度{A_i_1，…，A_i_n}，确定该成像周期T_img_i的单帧等效空间角度A_eq_i，并且，根据电子防抖权重Weis在该成像周期T_img_i内的权重值Weis_i、以及该成像周期T_img_i的单帧等效空间角度A_eq_i，确定第一电子防抖补偿角度Aeis_i，以根据该第一电子防抖补偿角度Aeis_i对该成像周期T_img_i结束时的刷新时刻得到的图像30实施图像处理，使该图像产生用于实现电子防抖的像素阵列调节。
针对实时偏摆空间角度{Yaw_i_1，…，Yaw_i_n}校正后的实时偏摆相对角度Yaw_ois_i_j＝Yaw_i_j×Wois_yaw_i/(Wois_yaw_i+Weis_yaw_i)，其表示为抵消摄像机抖动时的实时偏摆空间角度{Yaw_i_1，…，Yaw_i_n}所引发的成像坐标系相对于空间偏摆基准角度Yaw_abs_ref的偏摆偏差，而使成像坐标系相对于相对偏摆基准角度Yaw_rlv_ref(例如镜头光轴在机身坐标系中的偏摆方向上的装配角度)的角度值；Wois_yaw_i表示对应实时偏摆空间角度Yaw_i_j的光学偏摆防抖权重，Weis_yaw_i表示对应实时偏摆空间角度Yaw_i_j的电子偏摆防抖权重。
针对实时仰俯空间角度{Pitch_i_1，…，Pitch_i_n}校正后的实时仰俯相对角度Pitch_ois_i_j＝Pitch_i_j×Wois_pitch_i/(Wois_pitch_i+Weis_pitch_i)，其表示为抵消摄像机抖动时的实时仰俯空间角度{Pitch_i_1，…，Pitch_i_n}所引发的成像坐标系相对于空间仰俯基准角度Pitch_abs_ref的仰俯偏差，而使成像坐标系相对于相对仰俯基准角度Pitch_rlv_ref(例如镜头光轴在机身坐标系中的仰俯方向上的装配角度)的角度值；Wois_pitch_i表示对应实时仰俯空间角度Pitch_i_j的光学仰俯防抖权重，Weis_pitch_i表示对应实时仰俯空间角度Pitch_i_j的电子仰俯防抖权重。
假设针对实时翻转空间角度{Roll_i_1，…，Roll_i_n}也实施光学防抖和电子防抖的协同防抖， 则，针对实时翻转空间角度{Roll_i_1，…，Roll_i_n}校正后的实时仰俯翻转角度Roll_ois_i_j＝Roll_i_j×Wois_roll_i/(Wois_roll_i+Weis_roll_i)，其表示为抵消摄像机抖动时的实时翻转空间角度{Roll_i_1，…，Roll_i_n}所引发的成像坐标系相对于空间翻转基准角度Roll_abs_ref的翻转偏差，而使成像坐标系相对于相对翻转基准角度Roll_rlv_ref(例如镜头光轴在机身坐标系中的翻转方向上的装配角度)的角度值；Wois_roll_i表示对应实时翻转空间角度Roll_i_j的光学翻转防抖权重，Weis_roll_i表示对应实时翻转空间角度Roll_i_j的电子翻转防抖权重。
电子防抖实施图像处理而使像素阵列在水平方向上的平移，依据的是电子防抖偏摆补偿角度Yaw_eis_i＝Yaw_eq_i×Weis_yaw_i/(Wois_yaw_i+Weis_yaw_i)；
电子防抖实施图像处理而使像素阵列在竖直方向上的平移，依据的是电子防抖仰俯补偿角度Pitch_eis_i＝Pitch_eq_i×Weis_pitch_i/(Wois_pitch_i+Weis_pitch_i)；
假设针对翻转抖动也实施光学防抖和电子防抖的协同防抖，则，电子防抖实施图像处理而使像素阵列在翻转方向上的平转，依据的是电子防抖翻转补偿角度Roll_eis_i＝Roll_eq_i×Weis_roll_i/(Wois_roll_i+Weis_roll_i)；若针对翻转空间角度{Roll_i_1，…，Roll_i_n}的防抖完全依靠电子防抖，则，电子防抖根据单帧等效翻转空间角度实施图像处理而使像素阵列在翻转方向上的平转，即，Roll_eis_i＝Roll_eq_i。
在该实施例中，处理器组件90对光学防抖权重Wois的权重值Wois_i与电子防抖权重Weis的权重值Weis_i的更新，是基于对光学防抖对成像坐标系的实时相对角度的校正能力、以及电子防抖对图像清晰度的容忍度的考量，这是因为：
光学防抖受限于光学镜组31和感光元件32之间的相对位姿的物理调节性能，若成像坐标系被校正至足以抵消实时空间角度{A_i_1，…，A_i_n}的实时相对角度所需的物理调节性能，超出光学镜组31和感光元件32之间的相对位姿的物理调节性能之外，则，容易引发光学防抖的失效。
电子防抖依赖于感光元件32得到的图像的清晰度，若成像坐标系在成像周期(尤其是成像周期内的曝光期间)由于过大的实时空间角度而产生较大震动，并导致图像的清晰度被降低至不足以通过图像处理而被改善的程度(例如模糊)，则，容易引发电子防抖的失效。
基于上述原因，在任意一个成像周期T_img_i，摄像机的实时空间角度{A_i_1，…，A_i_n}可以只有在摄像机的光学防抖能力范围内的一部分(即与校正后的实时相对角度{Aois_i_1，…，Aois_i_n}反向等值的部分)被光学防抖抵消，以助于降低光学防抖失效的风险；并且，基于光学防抖所抵消的一部分空间角度，还可以降低图像清晰度由于空间角度过大而过低的风险，从而，在该成像周期T_img_i结束时的刷新时刻，通过电子防抖对清晰度被光学防抖改善后的图像实施图像处理，可以补偿摄像机的实时空间角度{A_i_1，…，A_i_n}的剩余部分(即A_i_j的绝对值与Aois_i_j的绝对值的差值)对图像质量的影响。进而，通过以电子防抖补偿光学防抖的双模式结合，有助于降低防抖失效的风险。
优选地，光学防抖权重Wois的初始值Wois_0可以取1、电子防抖权重Weis的初始值Weis_0可以取0。即，处理器组件90可以进一步用于响应于摄像机的上电启动的完成，将光学防抖权重Wois初始赋值为1、并将电子防抖权重Weis初始赋值为0，以实现光学防抖优先的协同原则。也就是说，可以将光学防抖权重值初始赋值为1、并将电子防抖权重值初始赋值为0。
即，光学偏摆防抖权重Wois_yaw的初始值Wois_yaw_0可以取1、电子偏摆防抖权重Weis_yaw的初始值Weis_yaw_0可以取0；光学俯仰防抖权重Wois_pitch的初始值Wois_pitch_0可以取1、电子俯仰防抖权重Weis_pitch的初始值Weis_pitch_0可以取0；光学翻转防抖权重Wois_roll的初始值Wois_roll_0可以取1、电子翻转防抖权重Weis_roll的初始值Weis_roll_0可以取0。
假设将感光元件32的任意一个成像周期T_img_i看作是第一成像周期，并且，假设：
光学防抖权重Wois在感光元件32的第一成像周期T_img_i内的权重值Wois_i，意图将实时相对角度{Aois_i_1，…，Aois_i_n}限制在摄像机的光学防抖能力范围内，以使得第一成像周期T_img_i内的实时相对角度{Aois_i_1，…，Aois_i_n}具有相对于实时空间角度{A_i_1，…，A_i_n}的实时角度差值{ΔA_i_1，…，ΔA_i_n}，其中，ΔA_i_j＝|A_i_j|-|Aois_i_j|，该实时角度差值ΔA_i_j可以看作是实时空间角度A_i_j中未被校正后的实时相对角度Aois_i_j完全抵消的差幅部分或残差部分。也就是说，可以根据第一成像周期内的光学防抖权重值对实时相对角度{Aois_i_1，…，Aois_i_n}进行角度校正，进而，使得角度校正后的实时相对角度{Aois_i_1，…，Aois_i_n}在摄像机的光学防抖能力范围内，且角度校正后的实时相对角度{Aois_i_1，…，Aois_i_n}具有相对于实时空间角度{A_i_1，…，A_i_n}的实时角度差值{ΔA_i_1，…，ΔA_i_n}。
图4为如图3所示协同防抖采用的电子防抖补偿光学防抖的示例性效果图。在图4中，示出了两条拟合曲线W51和W52，其中，拟合曲线W51表示实时相对角度{Aois_i_1，…，Aois_i_n}在光学偏摆防抖权重Wois_yaw取初始值1时的理想值，拟合曲线W52表示实时相对角度{Aois_i_1，…，Aois_i_n}在第一成像周期T_img_i内实际发生的真实值，故，拟合曲线W51可以称为实时相对角度{Aois_i_1，…，Aois_i_n}的理想拟合曲线，拟合曲线W52可以称为实时相对角度{Aois_i_1，…，Aois_i_n}的真实拟合曲线，并且，理想拟合曲线W51与实时空间角度{A_i_1，…，A_i_n}的拟合曲线W20以第一频率f_gyro的一个频率周期T_gyro的延迟反向等值。
实时角度差值{ΔA_i_1，…，ΔA_i_n}可以由理想拟合曲线W51与真实拟合曲线W52之间的偏差等效表示。
在此情况下，处理器组件90可以具体用于根据第一成像周期T_img_i内的实时空间角度{A_i_1，…，A_i_n}，确定该第一成像周期T_img_i内的第一单帧等效空间角度A_eq_i，并且，根据电子防抖权重Weis在该第一成像周期T_img_i内的权重值Weis_i、以及该第一单帧等效空间角度A_eq_i，确定第一电子防抖补偿角度Aeis_i＝A_eq_i×Weis_i/(Wois_i+Weis_i)，以根据该第一电子防抖补偿角度Aeis_i对该第一成像周期T_img_i结束时的第一刷新时刻得到的图像30实施图像处理，使该图像30产生用于补偿实时角度差值{ΔA_i_1，…，ΔA_i_n}的像素阵列调节，以实现电子防抖。也就是，以该方式实现的电子防抖可以认为是对光学防抖的补偿措施。
从而，基于光学防抖校正后的实时相对角度{Aois_i_1，…，Aois_i_n}、以及电子防抖对图像30的像素阵列调节的共同作用，光学防抖和电子防抖的协同防抖可以改善在摄像机抖动期间内拍摄到的图像质量。并且，相比于完全依赖光学防抖持续对成像坐标系的相对坐标实施满额校正的情况，协同防抖可以在实时空间角度{A_i_1，…，A_i_n}超出光学防抖能力范围的情况下，缓解光学防抖部分失效的风险；相比于完全依靠电子防抖的图像处理而改善图像质量的情况，协同防抖可以在实时空间角度{A_i_1，…，A_i_n}足以引发图像模糊的情况下，降低由于图像模糊而导致电子防抖失效的风险。
图5为适用于如图3所示协同防抖的权重更新过程的原理示意图。请参见图5，假设将第一成像周期T_img_i的前一成像周期T_img_i-1看作第二成像周期，则：
处理器组件90可以进一步用于根据该第二成像周期T_img_i-1内的实时空间角度{A_i-1_1，…，A_i-1_n}以及第二电子防抖补偿角度Aeis_i-1，确定光学防抖权重Wois在第一成像周期T_img_i内的权重值Weis_i和电子防抖权重Weis在第一成像周期T_img_i内的权重值Weis_i，以促使：
在第一成像周期T_img_i内，实时相对角度{A_i_1，…，A_i_n}被意图限制为收敛在预设的角度校正极限范围内，其中，该角度校正极限范围是根据光学防抖能力范围确定的；以及，
在第一成像周期T_img_i的第一刷新时刻得到的图像30的清晰度不低于预设的清晰度阈值，其中，该清晰度阈值是根据预先设定的用于避免电子防抖失效的最低图像清晰度确定的。
也就是说，基于上述处理可以确定出第一成像周期内的光学防抖权重值，并基于该光学防抖权重值对第一成像周期内的实时相对角度进行角度校正，角度校正后的实时相对角度收敛在预设的角度校正极限范围内。
优选地，处理器组件90可以进一步用于：
根据该第二成像周期T_img_i-1内的实时空间角度{A_i-1_1，…，A_i-1_n}，确定该第二成像周期T_img_i-1内的第二单帧等效空间角度A_eq_i-1，并且，确定第二单帧等效空间角度A_eq_i-1超出角度校正极限范围的幅度，作为光学超额幅度Aois_ex_i-1。光学超额幅度Aois_ex_i-1也可以称为第一幅度。第二单帧等效空间角度A_eq_i-1也就是与第二成像周期T_img_i-1关联的单帧等效空间角度。
其中，该光学超额幅度Aois_ex_i-1产生使电子防抖权重Weis的权重值增大的权重变化趋势，即，该光学超额幅度Aois_ex_i-1促使电子防抖权重Weis在第一成像周期T_img_i内的权重值Weis_i大于在第二成像周期T_img_i-1内的权重值Weis_i-1。也就是说，第一幅度的增大能够使电子防抖权重值产生增大的变化趋势。
以及，
确定第二成像周期T_img_i-1的第二电子防抖补偿角度Aeis_i-1相比于在第二成像周期T_img_i-1之前的第三成像周期T_img_i-2的第三电子防抖补偿角度Aeis_i-2的角度偏差，作为电子防抖补偿角度帧间偏差ΔAeis_i-1＝|Aeis_i-1-Aeis_i-2|，并且，确定电子防抖补偿角度帧间偏差ΔAeis_i-1超出预设的帧间抖幅阈值A_int_th的幅度，作为电子超额幅度Aeis_ex_i-1。电子超额幅度Aeis_ex_i-1也可以称为第二幅度。
其中，该电子超额幅度Aeis_ex_i-1可能引发图像30的清晰度降低，并且，该电子超额幅度Aeis_ex_i-1产生使光学防抖权重Wois的权重值增大的权重变化趋势，即，该电子超额幅度Aeis_ex_i-1促使光学防抖权重Wois在第一成像周期T_img_i内的权重值Wois_i大于在第二成像周期T_img_i-1内的权重值Wois_i-1。也就是说，第二幅度的增大能够使光学防抖权重值产生增大的变化趋势。
例如，处理器组件90可以进一步根据光学超额幅度Aois_ex_i-1在光学镜组31所在镜头10的视场角A_fov中的角度占比、以及电子防抖权重Weis在第二成像周期T_img_i-1内的权重值Weis_i-1，确定电子防抖权重Weis在第一成像周期T_img_i内的权重值Weis_i，并且，根据电子超额幅度Aeis_ex_i-1在光学镜组31所在镜头10的视场角A_fov中的角度占比、以及光学防抖权重Wois在第二成像周期T_img_i-1内的权重值Wois_i-1，确定光学防抖权重Wois在第一成像周期T_img_i内的权重值Wois_i。该确定方式可以表示为：
Wois_i＝Wois_i-1+Aeis_ex_i-1/A_fov；
Weis_i＝Weis_i-1+Aois_ex_i-1/A_fov。
针对偏摆抖动、仰俯抖动、以及翻转抖动，光学超额幅度Aois_ex_i-1可以分别包括光学偏转超额幅度Yaw_ois_ex_i-1、光学仰俯超额幅度Pitch_ois_ex_i-1以及光学翻转超额幅度Roll_ois_ex_i-1，电子超额幅度Aeis_ex_i-1可以具体包括电子偏转超额幅度Yaw_eis_ex_i-1、电子仰俯超额幅度Pitch_eis_ex_i-1以及电子翻转超额幅度Roll_eis_ex_i-1，并且，上述的确定方式可以具体表示为：
Wois_yaw_i＝Wois_yaw_i-1+Yaw_eis_ex_i-1/A_fov_yaw；
Weis_yaw_i＝Weis_yaw_i-1+Yaw_ois_ex_i-1/A_fov_yaw；
Wois_pitch_i＝Wois_pitch_i-1+Pitch_eis_ex_i-1/A_fov_pitch；
Weis_pitch_i＝Weis_pitch_i-1+Pitch_ois_ex_i-1/A_fov_pitch；
Wois_roll_i＝Wois_roll_i-1+Roll_eis_ex_i-1/A_fov_roll；
Weis_roll_i＝Weis_roll_i-1+Roll_ois_ex_i-1/A_fov_roll；
其中，A_fov_yaw表示视场角A_fov中的水平视场角，A_fov_pitch表示视场角中的垂直视场角，A_fov_roll表示视场角A_fov的角度范围(例如360°)。
如前文所述，防抖驱动组件70可以通过调节光学镜组31与感光元件32之间的相对位姿，实现对成像坐标系的持续角度校正，在此情况下，光学防抖能力范围可以是根据光学镜组31与感光元件32之间的相对位姿的物理调节性能确定的。其中，光学镜组31与感光元件32之间的相对位姿的物理调节性能可以涉及相对位姿的调节速率以及位姿极限中的至少之一。也就是说，防抖驱动组件70可以通过调节光学镜组31与感光元件32之间的相对位姿，实现对实时相对角度的持续地角度校正。
在光学镜组31与感光元件32之间的相对位姿的物理调节性能可以涉及相对位姿的调节速率的情况下，处理器组件90可以进一步用于确定第二单帧等效空间角度A_eq_i-1所需的目标速度Vobj_i-1超出光学镜组31与感光元件32之间的相对位姿调节的预设最大速度Vmax的速度差(即Vobj_i-1-Vmax)，并且，光学超额幅度Aois_ex_i-1可以包括第二单帧等效空间角度A_eq_i-1中引发该速度差的幅度，可以称为速度超限幅度Aois_v_ex_i-1。
其中，该目标速度Vobj_i-1表示：为使成像坐标系从当前相对角度Aois_i-1_n校正至与第二单帧等效空间角度A_eq_i-1反向等幅的目标速度理想校正角度Aois_v_obj_i-1，光学镜组31与感光元件32之间的相对位姿在第一频率f_gyro的一个频率周期T_gyro内的理论变化速率。
也就是，目标速度Vobj_i-1×T_gyro×Kapm＝|Aois_v_obj_i-1-Aois_i-1_n|，Kapm为预先设定的转换系数，该转换系数Kapm是根据光学镜组31与感光元件32之间的相对位姿的物理调节性能确定的，表示光学镜组31与感光元件32之间的相对位姿每调节预设的单位步长所能够产生的成像坐标系的单位校正幅度。
并且，预设最大速度Vmax是根据光学镜组31与感光元件32之间的相对位姿的物理调节性能所能支持的极限速度确定的。
若该目标速度Vobj_i-1表示对应偏摆方向的目标水平速度v_yaw_obj_i-1，预设最大速度Vmax表示光学镜组31与感光元件32之间的相对位姿调节的最大水平速度v_yaw_max，并且，目标速度理想校正角度Aois_v_obj_i-1表示与第二单帧等效空间角度A_eq_i-1所表示的单帧偏摆等效空间角度Yaw_eq_i-1反向等幅的目标水平速度理想校正角度Yaw_ois_v_obj_i-1，则：
v_yaw_obj_i-1×T_gyro×Kapm_yaw＝|Yaw_ois_v_obj_i-1-Yaw_ois_i-1_n|，其中，Kapm_yaw为预先设定的偏摆转换系数，该偏摆转换系数Kapm_yaw是根据光学镜组31与感光元件32之间的水平相对位置的物理调节性能确定的，表示光学镜组31与感光元件32之间的水平相对位置每调节预设的水平距离单位步长所能够产生的成像坐标系的单位偏摆校正幅度。
若该目标速度Vobj_i-1表示对应仰俯方向的目标竖直速度v_pitch_obj_i-1，预设最大速度Vmax表示光学镜组31与感光元件32之间的相对位姿调节的最大竖直速度v_pitch_max，并且，目标速度理想校正角度Aois_v_obj_i-1表示与第二单帧等效空间角度A_eq_i-1所表示的单帧仰俯等效空间角度Pitch_eq_i-1反向等幅的目标竖直速度理想校正角度Pitch_ois_v_obj_i-1，则：
v_pitch_obj_i-1×T_gyro×Kapm_pitch＝|Pitch_ois_obj_i-1-Pitch_ois_i-1_n|，其中，Kapm_pitch为预先设定的仰俯转换系数，该仰俯转换系数Kapm_pitch是根据光学镜组31与感光元件32之间的竖直相对位置的物理调节性能确定的，表示光学镜组31与感光元件32之间的竖直相对位置每调节预设的竖直距离单位步长所能够产生的成像坐标系的单位仰俯校正幅度。
若该目标速度Vobj_i-1表示对应翻转方向的目标角速度ω_roll_obj_i-1，预设最大速度Vmax 表示光学镜组31与感光元件32之间的相对位姿调节的最大角速度ω_roll_max，并且，目标速度理想校正角度Aois_v_obj_i-1表示与第二单帧等效空间角度A_eq_i-1所表示的单帧翻转等效空间角度Roll_eq_i-1反向等幅的目标角速度理想翻转校正角度Roll_ois_obj_i-1，则：
ω_roll_obj_i-1×T_gyro×Kapm_roll＝|Roll_ois_obj_i-1-Roll_ois_i-1_n|，其中，Kapm_roll为预先设定的翻转转换系数，该翻转转换系数Kapm_roll是根据光学镜组31与感光元件32之间的平转相对角度的物理调节性能确定的，表示光学镜组31与感光元件32之间的平转相对角度每调节预设的角度单位步长所能够产生的成像坐标系的单位翻转校正幅度。
图6为如图5所示权重更新过程中基于速度极限确定的光学超额幅度的示例性效果图。在图6中示出了表示目标速度理想校正角度Aois_v_obj_i-1的拟合曲线W71、以及表示速度超限幅度Aois_v_ex_i-1的拟合曲线W72，通过图6可以看出：
成像坐标系在一个频率周期T_gyro内，以目标速度Vobj_i-1从当前相对角度Aois_i-1_n校正至目标速度理想校正角度Aois_v_obj_i-1，该校正过程所需的目标速度理想校正行程可以表示为目标校正幅度ΔAois_v_obj_i-1＝|Aois_v_obj_i-1-Aois_i-1_n|；
若该目标速度理想校正行程ΔAois_v_obj_i-1超出了预设最大速度Vmax所能够支持的极限速度校正行程ΔAois_v_max＝Vmax×T_gyro×Kapm，则，第二单帧等效空间角度A_eq_i-1中引发该速度差(即，Vobj_i-1-Vmax)的速度超限幅度Aois_v_ex_i-1，可以包括目标速度理想校正行程ΔAois_v_obj_i-1与极限速度校正行程ΔAois_v_max的差值，即，Aois_v_ex_i-1＝ΔAois_v_obj_i-1-ΔAois_v_max；
若该速度理想校正行程ΔAois_v_obj_i-1未超出极限速度校正幅度ΔAois_v_max，则，第二单帧等效空间角度A_eq_i-1中引发速度差(即Vobj_i-1-Vmax)的速度超限幅度Aois_v_ex_i-1为0。
即，若第二单帧等效空间角度A_eq_i-1所需的目标速度Vobj_i-1超出光学镜组31与感光元件32之间的相对位姿调节的预设最大速度Vmax，则：
Aois_v_ex_i-1＝|Aois_v_obj_i-1-Aois_i-1_n|-Vmax×T_gyro×Kapm；
否则，Aois_v_ex_i-1＝0。
假设成像坐标系在第一频率f_gyro的一个频率周期T_gyro内，从当前偏摆相对角度Yaw_ois_i-1_n校正至目标水平速度理想校正角度Yaw_ois_v_obj_i-1的目标水平速度理想校正行程ΔYaw_ois_v_obj_i-1＝|Yaw_ois_v_obj_i-1-Yaw_ois_i-1_n|，并且，最大水平速度v_yaw_max所能够支持的极限水平速度校正行程ΔYaw_ois_v_max被确定为v_yaw_max×T_gyro×Kapm_yaw：
若该目标水平速度理想校正行程ΔYaw_ois_v_obj_i-1，超出了极限水平速度校正行程ΔYaw_ois_v_max，则，单帧偏摆等效空间角度Yaw_eq_i-1中引发水平速度差(即v_yaw_obj_i-1–v_yaw_max)的水平速度超限幅度Yaw_ois_v_ex_i-1，可以包括目标水平速度理想校正行程ΔYaw_ois_v_obj_i-1与极限水平速度校正行程ΔYaw_ois_v_max的差值，否则，偏摆速度超限幅度Yaw_ois_v_ex_i-1取0。
假设成像坐标系在第一频率f_gyro的一个频率周期T_gyro内，从当前仰俯相对角度Pitch_ois_i-1_n校正至目标竖直速度理想校正角度Pitch_ois_v_obj_i-1的目标竖直速度理想校正行程ΔPitch_ois_v_obj_i-1＝|Pitch_ois_v_obj_i-1-Pitch_ois_i-1_n|，并且，最大竖直速度v_pitch_max所能够支持的极限竖直速度校正行程ΔPitch_ois_v_max被确定为v_pitch_max×T_gyro×Kapm_pitch：
若该目标竖直速度理想校正行程ΔPitch_ois_v_obj_i-1，超出了极限竖直速度校正行程ΔPitch_ois_v_max，则，单帧仰俯等效空间角度Pitch_eq_i-1中引发竖直速度差(即v_pitch_obj_i-1–v_pitch_max)的竖直速度超限幅度Pitch_ois_v_ex_i-1，可以包括目标竖直速度理想校正行程Δ Pitch_ois_v_obj_i-1与极限竖直速度校正行程ΔPitch_ois_v_max的差值，否则，竖直速度超限幅度Pitch_ois_v_ex_i-1取0。
假设成像坐标系在第一频率f_gyro的一个频率周期T_gyro内，从当前翻转相对角度Roll_ois_i-1_n校正至目标角速度理想校正角度Roll_ois_ω_obj_i-1的目标角速度理想校正行程ΔRoll_ois_ω_obj_i-1＝|Roll_ois_ω_obj_i-1-Roll_ois_i-1_n|，最大角速度ω_roll_max支持的极限角速度校正行程ΔRoll_ois_ω_max被确定为ω_roll_max×T_gyro×Kapm_roll)：
若该目标角速度理想校正行程ΔRoll_ois_ω_obj_i-1，超出了上述的极限角速度校正行程ΔRoll_ois_ω_max，则，单帧翻转等效空间角度Roll_eq_i-1中引发平转角速度差(即ω_roll_obj_i-1–ω_roll_max)的角速度超限幅度Roll_ois_ω_ex_i-1，可以包括目标角速度理想校正行程ΔRoll_ois_ω_obj_i-1与极限角速度校正行程ΔRoll_ois_ω_max的差值，否则，角速度超限幅度Roll_ois_ω_ex_i-1取0。
在光学镜组31与感光元件32之间的相对位姿的物理调节性能可以涉及相对位姿的位姿极限的情况下，处理器组件90可以进一步用于确定第二单帧等效空间角度A_eq_i-1所需的目标位姿Pobj_i-1超出预设位姿范围Pmax之外的位姿差(即Pobj_i-1-Pmax)，并且，光学超额幅度Aois_ex_i-1可以包括第二单帧等效空间角度A_eq_i-1中引发该位姿差的幅度，可以称为位姿超限幅度Aois_p_ex_i-1。
其中，该目标位姿Pobj_i-1表示，为使成像坐标系达到与第二单帧等效空间角度A_eq_i-1反向等幅的目标位姿理想校正角度Aois_p_obj_i-1，光学镜组31与感光元件32之间的相对位姿经调节后的理论相对位姿。
并且，预设位姿范围Pmax是根据光学镜组31与感光元件32之间的相对位姿的物理调节性能所能支持的极限位姿确定的，例如，预设位姿范围Pmax可以小于物理调节性能所能支持的极限位姿的物理调节范围，以避免光学镜组31与感光元件32之间的相对位姿处于物理调节范围的边界处(处于边界处的期间内的成像质量相对低)。
具体地，该目标位姿Pobj_i-1可以表示对应偏摆方向的目标水平位置p_yaw_obj_i-1、或目标竖直位置p_pitch_obj_i-1、或目标翻转角度p_roll_obj_i-1，相应地，目标位姿理想校正角度Aois_p_obj_i-1可以表示：
与第二单帧等效空间角度A_eq_i-1所表示的单帧偏摆等效空间角度Yaw_eq_i-1反向等幅的目标水平位置理想校正角度Yaw_ois_p_obj_i-1；或者，
与第二单帧等效空间角度A_eq_i-1所表示的单帧仰俯等效空间角度Pitch_eq_i-1反向等幅的目标竖直位置理想校正角度Pitch_ois_p_obj_i-1；或者，
与第二单帧等效空间角度A_eq_i-1所表示的单帧翻转等效空间角度Roll_eq_i-1反向等幅的目标翻转角度理想校正角度Roll_ois_p_obj_i-1。
图7为如图5所示权重更新过程中基于位姿极限确定的光学超额幅度的示例性效果图。在图7中示出的拟合折线W80表示目标位姿校正超幅Aois_p_ex_i-1，该目标位姿校正超幅Aois_p_ex_i-1是目标位姿理想校正角度Aois_p_obj_i-1(例如2°)超出预设位姿范围Pmax所能支持的极限校正角度Aois_p_max(例如1.8°)的部分。
即，若第二单帧等效空间角度A_eq_i-1所需的目标位姿Pobj_i-1超出预设位姿范围Pmax，则：
Aois_p_ex_i-1＝Aois_p_obj_i-1-Aois_p_max；
否则，Aois_p_ex_i-1＝0。
假设目标水平位置p_yaw_obj_i-1超出了预设位姿范围Pmax中的极限水平位置p_yaw_max，导致成像坐标系的目标水平位置理想校正角度Yaw_ois_p_obj_i-1，超出了预设位姿范围Pmax所能 支持的极限偏摆校正角度Yaw_ois_p_max，则，单帧偏摆等效空间角度Yaw_eq_i-1中引发水平位置差(即p_yaw_obj_i-1–p_yaw_max)的水平位置超限幅度Yaw_ois_p_ex_i-1＝Yaw_ois_p_obj_i-1-Yaw_ois_p_max，否则，水平位置超限幅度Yaw_ois_p_ex_i-1取0。
假设目标竖直位置p_pitch_obj_i-1超出了预设位姿范围Pmax中的极限竖直位置p_pitch_max，导致成像坐标系的目标竖直位置理想校正角度Pitch_ois_p_obj_i-1，超出了预设位姿范围Pmax所能支持的极限仰俯校正角度Pitch_ois_p_max，则，单帧仰俯等效空间角度Pitch_eq_i-1中引发竖直位置差(即p_pitch_obj_i-1–p_pitch_max)的竖直位置超限幅度Pitch_ois_p_ex_i-1＝Pitch_ois_p_obj_i-1-Pitch_ois_p_max，否则，竖直位置超限幅度Pitch_ois_p_ex_i-1取0。
假设目标翻转角度p_roll_obj_i-1超出了预设位姿范围Pmax中的极限翻转角度p_roll_max，导致成像坐标系的目标翻转角度理想校正角度Roll_ois_p_obj_i-1，超出了预设位姿范围Pmax所能支持的极限翻转校正角度Roll_ois_p_max(该极限翻转校正角度Roll_ois_p_max通常小于±180°)，则，单帧翻转等效空间角度Roll_eq_i-1中引发翻转角度差(即p_roll_obj_i-1–p_roll_max)的翻转角度超限幅度Roll_ois_p_ex_i-1＝Roll_ois_p_obj_i-1-Roll_ois_p_max，否则，翻转角度超限幅度Roll_ois_p_ex_i-1取0。
上述如图6所示的基于速度极限确定的光学超额幅度Aois_ex_i-1＝Aois_v_ex_i-1，上述如图7所示的基于位姿极限确定的光学超额幅度Aois_ex_i-1＝Aois_p_ex_i-1，这两种方案可以彼此独立地分开使用。
或者，作为进一步优化，这两种方案也可以组合使用，在此情况下，光学超额幅度Aois_ex_i-1＝Aois_v_ex_i-1+Aois_p_ex_i-1，相应地，电子防抖权重Weis在第一成像周期T_img_i的权重值Weis_i被更新为：
Weis_i＝Weis_i-1+(Aois_v_ex_i-1+Aois_p_ex_i-1)/A_fov，即：
Weis_yaw_i＝Weis_yaw_i-1+(Yaw_ois_v_ex_i-1+Yaw_ois_p_ex_i-1)/A_fov_yaw；
Weis_pitch_i＝Weis_pitch_i-1+(Pitch_ois_v_ex_i-1+Pitch_ois_p_ex_i-1)/A_fov_pitch；
Weis_roll_i＝Weis_roll_i-1+(Roll_ois_v_ex_i-1+Roll_ois_p_ex_i-1)/A_fov_roll。
若Aois_v_ex_i-1和Aois_p_ex_i-1均为0，则，电子防抖权重Weis在第一成像周期T_img_i的权重值Weois_i与在第二成像周期T_img_i-1的权重值Weis_i-1相同。
对于电子超额幅度Aeis_ex_i-1的确定，如前文所述，处理器组件90用于确定第二成像周期T_img_i-1相比于之前的第三成像周期T_img_i-2的电子防抖补偿角度帧间偏差ΔAeis_i-1(即|Aeis_i-1-Aeis_i-2|)超出预设的帧间抖幅阈值A_int_th的电子超额幅度Aeis_ex_i-1，即：
第二成像周期T_img_i-1相比于之前的第三成像周期T_img_i-2的电子防抖偏摆补偿角度帧间偏差ΔYaw_eis_i-1(即|Yaw_eis_i-1-Yaw_eis_i-2|)超出帧间偏摆抖幅阈值Yaw_int_th的偏摆抖幅超限幅度Yaw_eis_ex_i-1＝ΔYaw_eis_i-1-ΔYaw_int_th；
第二成像周期T_img_i-1相比于之前的第三成像周期T_img_i-2的电子防抖仰俯补偿角度帧间偏差ΔPitch_eis_i-1(即|Pitch_eis_i-1-Pitch_eis_i-2|)超出帧间仰俯抖幅阈值Pitch_int_th的仰俯抖幅超限幅度Pitch_eis_ex_i-1＝ΔPitch_eis_i-1-ΔPitch_int_th；
第二成像周期T_img_i-1相比于之前的第三成像周期T_img_i-2的电子防抖翻转补偿角度帧间偏差ΔRoll_eis_i-1(即|Roll_eis_i-1-Roll_eis_i-2|)超出帧间翻转抖幅阈值Roll_int_th的翻转抖幅超限幅度Roll_eis_ex_i-1＝ΔRoll_eis_i-1-ΔRoll_int_th。
除此之外，处理器组件90还可以进一步用于获取与光学镜组31所在镜头10的快门时间t_sh关联变化的角度阈值，可以称为防模糊角度阈值A_ed(sh)，用作确定电子超额幅度Aeis_ex_i-1的帧间抖幅阈值A_int_th，即，该帧间抖幅阈值A_int_th可以包括获取到的防模糊角度阈值A_ed(sh)， 并且，该防模糊角度阈值A_ed(sh)是根据与快门时间t_sh关联的清晰度阈值确定的。
若电子防抖补偿角度帧间偏差ΔAeis_i-1超出预设的帧间抖幅阈值A_int_th，即超出防模糊角度阈值A_ed(sh)，则，可能引发图像30的清晰度低于预设清晰度阈值的电子超额幅度Aeis_ex_i-1＝ΔAeis_i-1-A_ed(sh)，进而，光学防抖权重Wois在第一成像周期T_img_i的权重值Wois_i可以被更新为：
Wois_i＝Wois_i-1+(ΔAeis_i-1-A_ed(sh))/A_fov；
否则，Aeis_ex_i-1＝0，光学防抖权重Wois在第一成像周期T_img_i的权重值Wois_i与在第二成像周期T_img_i-1的权重值Wois_i-1相同。
该防模糊角度阈值A_ed(sh)可以表示偏摆方向上的防模糊偏摆角度阈值Yaw_ed(sh)、或防模糊仰俯角度阈值Pitch_ed(sh)、或防模糊翻转角度阈值Roll_ed(sh)。
图8为如图5所示权重更新过程中关联快门时间确定的电子超额幅度的示例性效果图。在图8中，折线W91表示第二成像周期T_img_i-1的第二电子防抖补偿角度Aeis_i-1和第三成像周期T_img_i-2的第三电子防抖补偿角度Aeis_i-2，折线W92表示电子防抖补偿角度帧间偏差ΔAeis_i-1，折线W90则表示电子防抖补偿角度帧间偏差ΔAeis_i-1超出防模糊角度阈值A_ed(sh)的电子超额幅度Aeis_ex_i-1。
以上是对光学防抖权重Wois在第一成像周期T_img_i内的权重值Weis_i和电子防抖权重Weis在第一成像周期T_img_i内的权重值Weis_i的确定方式的举例说明。按照与第一成像周期T_img_i相同的原理，处理器组件90还可以进一步用于确定光学防抖权重Wois在跟随于第一成像周期T_img_i的第四成像周期T_img_i+1内的权重值Weis_i+1和电子防抖权重Weis在该第四成像周期T_img_i+1内的权重值Weis_i+1，本文对此不再赘述。
图9为本申请的另一个实施例中防抖的控制方法的示例性流程示意图。请参见图9，在另一个实施例中提供的该控制方法可以由摄像机的处理器组件以第二频率(例如感光元件产生图像的周期频率f_img)周期性地执行，并且，该控制方法可以包括：
S900：确定用于实现协同防抖的光学防抖权重和电子防抖权重在感光元件的第一成像周期内的权重值，该光学防抖权重在摄像机的感光元件的第一成像周期内的权重值，意图将感光元件在摄像机的机身坐标系中的实时相对角度，限制在摄像机的光学防抖能力范围内，以使得第一成像周期内的实时相对角度具有相对于摄像机在空间坐标系中的实时空间角度的实时角度差值。
例如，摄像机在空间坐标系中的实时空间角度，可以是根据诸如陀螺仪等抖动传感器以第一频率f_gyro持续产生的实时抖动数据确定的，并且，抖动传感器产生实时抖动数据的第一频率f_gyro，不低于感光元件成像的第二频率f_img的两倍。
S910：根据第一成像周期内的实时空间角度，确定第一成像周期的第一单帧等效空间角度。
例如，第一单帧等效空间角度可以取第一成像周期内的实时空间角度中的峰值角度。
S930：根据述电子防抖权重在所述第一成像周期内的权重值、以及第一单帧等效空间角度，确定第一电子防抖补偿角度，以根据确定的第一电子防抖补偿角度对第一成像周期结束时的第一刷新时刻得到的图像实施图像处理，通过使该图像产生用于补偿实时角度差值的像素阵列调节实现电子防抖。
本步骤之后，即可将下一成像周期作为第一成像周期、并返回S900继续执行上述流程。
基于上述流程，在感光元件的任意一个刷新周内，摄像机的实时空间角度可以只有在摄像机的光学防抖能力范围内的一部分被光学防抖抵消，以助于降低光学防抖失效的风险；并且，基于光学防抖所抵消的一部分空间角度，还可以降低图像清晰度由于空间角度过大而过低的风险，从而，在该成像周期结束时的刷新时刻，通过电子防抖对具有清晰度被光学防抖改善后的图像实施图像处理， 可以补偿摄像机的实时空间角度的剩余部分对图像质量的影响。进而，通过以电子防抖补偿光学防抖的双模式结合，有助于降低防抖失效的风险。
优选地，上述流程中的S900可以发生在第一成像周期之前，并且，可以具体包括：
根据第一成像周期之前的第二成像周期内的实时空间角度和第二电子防抖补偿角度，确定光学防抖权重和电子防抖权重在第一成像周期内的权重值，以促使：
在第一成像周期内，实时相对角度被意图限制为收敛在预设的角度校正极限范围内，该角度校正极限范围可以是根据所述光学防抖能力范围确定的；以及，
在第一刷新时刻得到的图像的清晰度不低于预设的清晰度阈值，该清晰度阈值可以是根据预先设定的用于避免电子防抖失效的最低图像清晰度确定的。
进一步地，为了确定光学防抖权重和电子防抖权重在第一成像周期内的权重值，可以具体采用如下方式：
根据第二成像周期内的实时空间角度，确定第二成像周期的第二单帧等效空间角度，并且，确定第二单帧等效空间角度超出角度校正极限范围的光学超额幅度，其中，光学超额幅度产生使电子防抖权重的权重值增大的权重变化趋势；
确定第二电子防抖补偿角度相比于在第二成像周期之前的第三成像周期的第三电子防抖补偿角度的电子防抖补偿角度帧间偏差，并且，确定电子防抖补偿角度帧间偏差超出预设的帧间抖幅阈值的电子超额幅度，其中，电子超额幅度引发图像的清晰度低于清晰度阈值，并且，电子超额幅度产生使光学防抖权重的权重值增大的权重变化趋势。
例如，若光学防抖对成像坐标系的实时相对角度的持续校正，通过调节摄像机的光学镜组与感光元件之间的相对位姿实现，并且，光学防抖能力范围是根据光学镜组与感光元件之间的相对位姿的物理调节性能确定的，则，在确定第二单帧等效空间角度超出角度校正极限范围的光学超额幅度时：
可以确定第二单帧等效空间角度所需的目标速度超出预设最大速度的速度差，其中，光学超额幅度可以包括第二单帧等效空间角度中引发该速度差的速度超限幅度；该目标速度表示，为使成像坐标系从当前相对角度校正至与第二单帧等效空间角度反向等幅的目标速度理想校正角度，光学镜组与感光元件之间的相对位姿的理论变化速率(例如在第一频率的一个频率周期内的理论变化速率)；并且，预设最大速度可以是根据物理调节性能所能支持的极限速度确定的；和/或，
可以确定第二单帧等效空间角度所需的目标位姿超出预设位姿范围之外的位姿差，其中，光学超额幅度包括第二单帧等效空间角度中引发该位姿差的位姿超限幅度；该目标位姿表示，为使成像坐标系达到与第二单帧等效空间角度反向等幅的目标位姿理想校正角度，相对位姿经调节后的理论相对位姿；预设位姿范围可以是根据物理调节性能所能支持的极限位姿确定的。
确定光学超额幅度和电子超额幅度的具体示例，可以参见前述实施例，此处不再赘述。
另外，该控制方法还可以进一步包括：获取与光学镜组所在镜头的快门时间关联变化的防模糊角度阈值，其中，确定电子超额幅度所使用的帧间抖幅阈值可以包括获取到的所述防模糊角度阈值，并且，该防模糊角度阈值可以是根据与所快门时间关联的清晰度阈值确定的。
另一方面，提供了一种控制设备，包括：处理器；用于存储处理器可执行指令的存储器；其中，所述处理器被配置为执行上述任一所述的控制方法的步骤。
另一方面，提供了一种计算机可读存储介质，所述计算机可读存储介质上存储有指令，所述指令被处理器执行时实现上述任一所述的控制方法的步骤。
另一方面，提供了一种包含指令的计算机程序产品，当其在处理器上运行时，使得处理器执行上述任一所述的控制方法的步骤。
以上所述仅为本申请的较佳实施例，并不用以限制本申请，凡在本申请的精神和原则之内，所做的任何修改、等同替换、改进等，均应包含在本申请保护的范围之内。

Claims (16)

  1. 一种摄像机，包括：
    抖动传感器，用于以第一频率持续产生所述摄像机的实时抖动数据，其中，所述第一频率不低于第二频率的两倍，所述第二频率为所述摄像机的感光元件成像的频率，并且，所述实时抖动数据用于确定所述摄像机在空间坐标系中的实时空间角度；
    处理器组件，用于以所述第二频率更新光学防抖权重值和电子防抖权重值；
    防抖驱动组件，用于根据所述光学防抖权重值，对所述感光元件的成像坐标系在所述摄像机的机身坐标系中的实时相对角度，实施与所述实时空间角度反向的持续地角度校正，以实现响应于所述实时空间角度的光学防抖；
    其中，根据所述感光元件的第一成像周期内的光学防抖权重值对实时相对角度进行角度校正，使得角度校正后的实时相对角度在所述摄像机的光学防抖能力范围内，以使得角度校正后的实时相对角度具有相对于实时空间角度的实时角度差值；
    并且，处理器组件还用于根据所述第一成像周期内的实时空间角度，确定所述第一成像周期的等效空间角度，作为第一等效空间角度，并且，根据所述第一成像周期内的电子防抖权重值、以及所述第一等效空间角度，确定第一电子防抖补偿角度，以根据所述第一电子防抖补偿角度调节第一刷新时刻得到的图像的像素阵列，得到像素阵列的调节幅度，所述像素阵列的调节幅度用于补偿所述实时角度差值，实现电子防抖；所述第一刷新时刻表示所述第一成像周期的结束时刻。
  2. 根据权利要求1所述的摄像机，其中，
    所述处理器组件用于根据所述第一成像周期之前的第二成像周期内的实时空间角度和第二电子防抖补偿角度，确定所述第一成像周期内的光学防抖权重值和电子防抖权重值，以促使：
    所述第一成像周期内角度校正后的实时相对角度收敛在预设的角度校正极限范围内；以及，
    在所述第一刷新时刻得到的图像的清晰度不低于预设的清晰度阈值；
    其中，所述角度校正极限范围是根据所述光学防抖能力范围确定的，并且，所述清晰度阈值是根据预先设定的用于避免所述电子防抖失效的最低图像清晰度确定的。
  3. 根据权利要求2所述的摄像机，其中，
    所述处理器组件进一步用于根据所述第二成像周期内的实时空间角度，确定所述第二成像周期的等效空间角度，作为第二等效空间角度，并且，确定所述第二等效空间角度超出所述角度校正极限范围的幅度，作为第一幅度，其中，所述第一幅度的增大使所述电子防抖权重值产生增大的变化趋势；
    所述处理器组件进一步用于确定所述第二电子防抖补偿角度相比于第三电子防抖补偿角度的角度偏差，作为第一角度偏差，所述第三电子防抖补偿角度表示在所述第二成像周期之前的第三成像周期的电子防抖补偿角度；并且，确定所述第一角度偏差超出预设的帧间抖幅阈值的幅度，作为第二幅度，其中，所述第二幅度使所述图像的清晰度低于所述清晰度阈值，并且，所述第二幅度的增大使所述光学防抖权重值产生增大的变化趋势。
  4. 根据权利要求3所述的摄像机，其中，
    所述处理器组件进一步根据所述第一幅度在所述摄像机光学镜组所在镜头的视场角中的角度占比、以及所述第二成像周期内的电子防抖权重值，确定所述第一成像周期内的电子防抖权重值；并且，根据所述第二幅度在所述视场角中的角度占比、以及所述第二成像周期内的光学防抖权重值，确定所述第一成像周期内的光学防抖权重值。
  5. 根据权利要求3所述的摄像机，其中，
    所述防抖驱动组件进一步用于通过调节所述摄像机的光学镜组与所述感光元件之间的相对位姿，实现对所述实时相对角度的所述持续地角度校正；
    其中，所述光学防抖能力范围是根据所述光学镜组与所述感光元件之间的相对位姿的物理调节性能确定的。
  6. 根据权利要求5所述的摄像机，其中，
    所述处理器组件进一步用于确定所述第二等效空间角度所需的目标速度超出预设最大速度的速度差，其中：
    所述第一幅度包括所述第二等效空间角度中引发所述速度差的幅度；
    所述目标速度表示，为使所述成像坐标系从当前相对角度校正至与所述第二等效空间角度反向等幅的目标速度理想校正角度，所述相对位姿在所述第一频率的一个频率周期内的理论变化速率；
    所述预设最大速度是根据所述物理调节性能所能支持的极限速度确定的。
  7. 根据权利要求5所述的摄像机，其中，
    所述处理器组件进一步用于确定所述第二等效空间角度所需的目标位姿超出预设位姿范围之外的位姿差，其中：
    所述第一幅度包括所述第二等效空间角度中引发所述位姿差的幅度；
    所述目标位姿表示，为使所述成像坐标系达到与所述第二等效空间角度反向等幅的目标位姿理想校正角度，所述相对位姿经调节后的理论相对位姿；
    所述预设位姿范围是根据所述物理调节性能所能支持的极限位姿确定的。
  8. 根据权利要求3所述的摄像机，其中，
    所述处理器组件进一步用于获取与所述光学镜组所在镜头的快门时间关联变化的角度阈值；
    其中，所述帧间抖幅阈值包括获取到的所述角度阈值，并且，所述角度阈值是根据与所述快门时间关联的所述清晰度阈值确定的。
  9. 根据权利要求1所述的摄像机，其中，所述处理器组件进一步用于响应于所述摄像机的上电启动完成，将所述光学防抖权重值初始赋值为1、并将所述电子防抖权重值初始赋值为0。
  10. 根据权利要求1至9中任一项所述的摄像机，其中，所述处理器组件进一步用于将所述感光元件的每个成像周期内的实时空间角度中的峰值角度确定为该成像周期内的等效空间角度。
  11. 根据权利要求1至9中任一项所述的摄像机，其中，
    所述摄像机的抖动包括在水平摆动方向上的偏摆抖动、在仰俯摆动方向上的仰俯抖动、以及在绕镜头光轴的旋转方向上的翻转抖动；
    所述抖动传感器产生所述实时抖动数据、所述处理器组件对所述光学防抖权重值和所述电子防抖权重值更新、所述防抖驱动组件实施的所述持续地角度校正、所述处理器组件通过图像处理对所述像素阵列的调节，是响应于所述偏摆抖动、所述仰俯抖动以及所述翻转抖动中的至少之一而被引发的。
  12. 一种防抖的控制方法，包括：
    确定摄像机的感光元件的第一成像周期内的光学防抖权重值和电子防抖权重值，其中，所述第一成像周期内的光学防抖权重值，用于对所述感光元件的成像坐标系在所述摄像机的机身坐标系中的实时相对角度，实施与所述摄像机在空间坐标系中的实时空间角度反向的持续地角度校正，使得角度校正后的实时相对角度在所述摄像机的光学防抖能力范围内，以使得角度校正后的实时相对角度具有相对于实时空间角度的实时角度差值；
    根据所述第一成像周期内的实时空间角度，确定所述第一成像周期的等效空间角度，作为第一等效空间角度；以及，
    根据所述第一成像周期内的电子防抖权重值、以及所述第一等效空间角度，确定第一电子防抖补偿角度，以根据所述第一电子防抖补偿角度调节第一刷新时刻得到的图像的像素阵列，得到像素阵列的调节幅度，所述像素阵列的调节幅度用于补偿所述实时角度差值，实现电子防抖。
  13. 根据权利要求12所述的控制方法，其中，确定摄像机的感光元件的第一成像周期内的光学防抖权重值和电子防抖权重值，包括：
    根据所述第一成像周期之前的第二成像周期内的实时空间角度和第二电子防抖补偿角度，确定所述第一成像周期内的光学防抖权重值和电子防抖权重值，以促使：
    所述第一成像周期内角度校正后的实时相对角度收敛在预设的角度校正极限范围内；以及，
    在所述第一刷新时刻得到的图像的清晰度不低于预设的清晰度阈值；
    其中，所述角度校正极限范围是根据所述光学防抖能力范围确定的，并且，所述清晰度阈值是根据预先设定的用于避免所述电子防抖失效的最低图像清晰度确定的。
  14. 根据权利要求13所述的控制方法，其中，确定所述第一成像周期内的光学防抖权重值和电子防抖权重值，包括：
    根据所述第二成像周期内的实时空间角度，确定所述第二成像周期的等效空间角度，作为第二等效空间角度，并且，确定所述第二等效空间角度超出所述角度校正极限范围的幅度，作为第一幅度，其中，所述第一幅度的增大使所述电子防抖权重值产生增大的变化趋势；
    确定所述第二电子防抖补偿角度相比于第三电子防抖补偿角度的角度偏差，作为第一角度偏差；所述第三电子防抖补偿角度表示在所述第二成像周期之前的第三成像周期的电子防抖补偿角度，并且，确定所述第一角度偏差超出预设的帧间抖幅阈值的幅度，作为第二幅度，其中，所述第二幅度使所述图像的清晰度低于所述清晰度阈值，并且，所述第二幅度的增大使所述光学防抖权重值产生增大的权化趋势。
  15. 根据权利要求14所述的控制方法，其中，还包括：
    根据所述第一幅度在所述摄像机的光学镜组所在镜头的视场角中的角度占比、以及所述第二成像周期内的电子防抖权重值，确定所述第一成像周期内的电子防抖权重值；并且，根据所述第二幅度在所述视场角中的角度占比、以及所述第二成像周期内的光学防抖权重值，确定所述第一成像周期内的光学防抖权重值。
  16. 根据权利要求14所述的控制方法，其中，
    对所述实时相对角度的角度校正，是通过调节所述摄像机的光学镜组与所述感光元件之间的相对位姿实现的；
    其中，所述光学防抖能力范围是根据所述光学镜组与所述感光元件之间的相对位姿的物理调节性能确定的；
    并且，确定所述第二等效空间角度超出所述角度校正极限范围的幅度，作为第一幅度，包括：
    确定所述第二等效空间角度所需的目标位姿超出预设位姿范围之外的位姿差，其中，所述第一幅度包括所述第二等效空间角度中引发所述位姿差的幅度；所述目标位姿表示，为使所述成像坐标系达到与所述第二等效空间角度反向等幅的目标位姿理想校正角度，所述相对位姿经调节后的理论相对位姿；所述预设位姿范围是根据所述物理调节性能所能支持的极限位姿确定的。
PCT/CN2022/105429 2021-07-21 2022-07-13 摄像机 Ceased WO2023001036A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202110822687.7A CN113452919B (zh) 2021-07-21 2021-07-21 用于利用光学防抖和电子防抖实现协同防抖的摄像机
CN202110822687.7 2021-07-21

Publications (1)

Publication Number Publication Date
WO2023001036A1 true WO2023001036A1 (zh) 2023-01-26

Family

ID=77816869

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2022/105429 Ceased WO2023001036A1 (zh) 2021-07-21 2022-07-13 摄像机

Country Status (2)

Country Link
CN (1) CN113452919B (zh)
WO (1) WO2023001036A1 (zh)

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118509707A (zh) * 2024-07-19 2024-08-16 四川新视创伟超高清科技有限公司 一种光学防抖与数字防抖融合方法及融合系统
CN119603557A (zh) * 2024-11-27 2025-03-11 维沃移动通信有限公司 防抖方法、装置、摄像模组、电子设备和可读存储介质

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN113452919B (zh) * 2021-07-21 2022-04-19 杭州海康威视数字技术股份有限公司 用于利用光学防抖和电子防抖实现协同防抖的摄像机
CN116055881B (zh) * 2022-07-15 2023-11-24 荣耀终端有限公司 一种光学防抖模块的控制方法及相关装置
CN116709023B (zh) * 2022-12-14 2024-03-26 荣耀终端有限公司 视频处理方法和装置

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170041545A1 (en) * 2015-08-06 2017-02-09 Invensense, Inc. Systems and methods for stabilizing images
US20170331998A1 (en) * 2016-05-11 2017-11-16 Htc Corporation Camera device, method for camera device, and non-transitory computer readable storage medium
US20190104255A1 (en) * 2017-10-03 2019-04-04 Google Llc Video Stabilization
CN109660718A (zh) * 2018-11-30 2019-04-19 Oppo广东移动通信有限公司 图像处理方法和装置、电子设备、计算机可读存储介质
CN110166697A (zh) * 2019-06-28 2019-08-23 Oppo广东移动通信有限公司 摄像头防抖方法、装置、电子设备和计算机可读存储介质
CN110278360A (zh) * 2019-06-28 2019-09-24 Oppo广东移动通信有限公司 图像处理方法和装置、电子设备、计算机可读存储介质
CN111711756A (zh) * 2020-06-28 2020-09-25 Oppo广东移动通信有限公司 一种图像防抖方法、电子设备及存储介质
CN113452919A (zh) * 2021-07-21 2021-09-28 杭州海康威视数字技术股份有限公司 用于利用光学防抖和电子防抖实现协同防抖的摄像机

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US11539887B2 (en) * 2018-03-23 2022-12-27 Huawei Technologies Co., Ltd. Video image anti-shake method and terminal
CN110266966A (zh) * 2019-06-28 2019-09-20 Oppo广东移动通信有限公司 图像生成方法和装置、电子设备、计算机可读存储介质
CN111246100B (zh) * 2020-01-20 2022-03-18 Oppo广东移动通信有限公司 防抖参数的标定方法、装置和电子设备

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20170041545A1 (en) * 2015-08-06 2017-02-09 Invensense, Inc. Systems and methods for stabilizing images
US20170331998A1 (en) * 2016-05-11 2017-11-16 Htc Corporation Camera device, method for camera device, and non-transitory computer readable storage medium
US20190104255A1 (en) * 2017-10-03 2019-04-04 Google Llc Video Stabilization
CN109660718A (zh) * 2018-11-30 2019-04-19 Oppo广东移动通信有限公司 图像处理方法和装置、电子设备、计算机可读存储介质
CN110166697A (zh) * 2019-06-28 2019-08-23 Oppo广东移动通信有限公司 摄像头防抖方法、装置、电子设备和计算机可读存储介质
CN110278360A (zh) * 2019-06-28 2019-09-24 Oppo广东移动通信有限公司 图像处理方法和装置、电子设备、计算机可读存储介质
CN111711756A (zh) * 2020-06-28 2020-09-25 Oppo广东移动通信有限公司 一种图像防抖方法、电子设备及存储介质
CN113452919A (zh) * 2021-07-21 2021-09-28 杭州海康威视数字技术股份有限公司 用于利用光学防抖和电子防抖实现协同防抖的摄像机

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
LEI QIONGYING, JIN WEIQI; GUO HONG; MI FENGWEN; ZHANG XU; HU LIANGLIANG: "Opto-mechatronics Joint Image Stabilization Based on MEMS Gyroscope and Piezoelectric Micro-pendulum Mirror", INFRARED TECHNOLOGY, KUNMING WULI YANJIUSUO, CN, vol. 40, no. 4, 16 October 2018 (2018-10-16), CN , pages 332 - 337, XP093026423, ISSN: 1001-8891 *

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN118509707A (zh) * 2024-07-19 2024-08-16 四川新视创伟超高清科技有限公司 一种光学防抖与数字防抖融合方法及融合系统
CN118509707B (zh) * 2024-07-19 2024-10-25 四川国创新视超高清视频科技有限公司 一种光学防抖与数字防抖融合方法及融合系统
CN119603557A (zh) * 2024-11-27 2025-03-11 维沃移动通信有限公司 防抖方法、装置、摄像模组、电子设备和可读存储介质

Also Published As

Publication number Publication date
CN113452919B (zh) 2022-04-19
CN113452919A (zh) 2021-09-28

Similar Documents

Publication Publication Date Title
CN113452919B (zh) 用于利用光学防抖和电子防抖实现协同防抖的摄像机
US10848676B2 (en) Semiconductor device and electronic device
US12063439B2 (en) Combined mechanical and electronic image stabilization
JP5011387B2 (ja) 撮像装置
US10348969B2 (en) Camera controller, image processing module, and semiconductor system
WO2019129061A1 (zh) 一种镜头模组和镜头模组的控制方法
JP7456043B2 (ja) 撮像支援装置、撮像支援システム、撮像システム、撮像支援方法、及びプログラム
JP2012037865A (ja) 像ブレ補正装置及びその制御方法及び撮像装置
CN104767930A (zh) 用于图像校正的设备和方法
JPH05199449A (ja) 防振カメラ
CN117641124A (zh) 图像稳定控制设备和方法以及存储介质
CN106303249A (zh) 视频防抖方法和设备
CN107040708A (zh) 图像稳定模块及相机模块
JPH0951469A (ja) 撮像装置及び画像揺れの補正方法
JP3460385B2 (ja) 手振れ補正装置
CN114761871A (zh) 摄像支援装置、摄像系统、摄像支援方法及程序
JP6912923B2 (ja) 撮像装置、撮像装置の駆動機構、およびそれらの制御方法
JP2025032618A (ja) 像ブレ補正の制御装置及び制御方法
JP6335671B2 (ja) 像振れ補正装置および像振れ補正方法
CN108234877A (zh) 一种空间光学稳像系统与方法
US20260025583A1 (en) Optical image stabilization system and method of controlling optical image stabilization
JP2003259194A (ja) 画像安定化装置
JP7336741B1 (ja) 撮像装置
EP4576800A1 (en) Method and apparatus in a camera system
CN121665111A (zh) 一种抖动校正方法、装置、存储介质和电子设备

Legal Events

Date Code Title Description
NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 22845201

Country of ref document: EP

Kind code of ref document: A1