WO2022178703A1 - 电子增稳方法及装置、可移动平台、成像装置 - Google Patents

电子增稳方法及装置、可移动平台、成像装置 Download PDF

Info

Publication number
WO2022178703A1
WO2022178703A1 PCT/CN2021/077620 CN2021077620W WO2022178703A1 WO 2022178703 A1 WO2022178703 A1 WO 2022178703A1 CN 2021077620 W CN2021077620 W CN 2021077620W WO 2022178703 A1 WO2022178703 A1 WO 2022178703A1
Authority
WO
WIPO (PCT)
Prior art keywords
attitude
imaging device
low
data
angular velocity
Prior art date
Application number
PCT/CN2021/077620
Other languages
English (en)
French (fr)
Inventor
罗玮
李兵
刘力源
Original Assignee
深圳市大疆创新科技有限公司
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 深圳市大疆创新科技有限公司 filed Critical 深圳市大疆创新科技有限公司
Priority to PCT/CN2021/077620 priority Critical patent/WO2022178703A1/zh
Publication of WO2022178703A1 publication Critical patent/WO2022178703A1/zh

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

Definitions

  • the present application relates to the technical field of image processing, and more particularly, to an electronic stabilization method, an electronic stabilization device, a movable platform, an imaging device, and a computer-readable storage medium.
  • the posture of the image can be adjusted, thereby eliminating the influence of the user's undesired motion (eg, shaking, etc.) on the posture of the image during the image capturing process.
  • the process of electronically stabilizing the image it is difficult to make the posture of the image after the electronically stabilizing process meet the user's expectation.
  • an electronic stabilization method In view of the above problems, an electronic stabilization method, electronic stabilization device, movable platform, imaging device, and computer-readable storage medium are proposed to overcome the above problems or at least partially solve the above problems.
  • a method for electronic stabilization of an image comprising: acquiring motion information of an imaging device at the current capture moment of the image and attitude data obtained based on measurement data of an attitude sensor, the The image is captured by the imaging device; target pose data for stabilizing the image is determined based at least in part on the motion information; obtained based on the target pose data and measurement data based on the pose sensor at the current capture moment Pose data processes the image.
  • an electronic stabilization device includes: a memory for storing executable instructions; a processor for executing the executable instructions stored in the memory Execute instructions to perform the following operations: acquiring motion information of the imaging device at the current capture moment of an image captured by the imaging device and attitude data obtained based on measurement data of the attitude sensor; based at least in part on the current motion
  • the target attitude data for stabilizing the image is determined based on the information; the image is processed according to the target attitude data and the attitude data obtained based on the measurement data of the attitude sensor at the current capture moment.
  • a movable platform carries an imaging device, the imaging device is used for capturing images; the movable platform further comprises: a memory for storing the movable platform execution instructions; a processor for executing the executable instructions stored in the memory to perform the following operations: acquiring motion information of the imaging device at the current capture moment of the image and attitude data obtained based on measurement data of the attitude sensor; Determining target pose data for stabilizing the image based at least in part on the motion information; and processing the image based on the target pose data and pose data obtained based on measurement data from a pose sensor at the current capture moment.
  • an imaging device includes an image sensor, and the image sensor is used for capturing images, wherein the imaging device further includes: a memory for storing a execution instructions; a processor for executing the executable instructions stored in the memory to perform the following operations: acquiring motion information of the imaging device at the current capture moment of the image and attitude data obtained based on measurement data of the attitude sensor; Determining target pose data for stabilizing the image based at least in part on the motion information; and processing the image based on the target pose data and pose data obtained based on measurement data of the pose sensor at the current capture moment.
  • a computer-readable storage medium storing executable instructions that, when executed by one or more processors, cause the one or more processors to process The controller performs the electronic stabilization method provided by the first aspect of the present application.
  • the motion information of the imaging device at the moment of image capture can reflect the image posture desired by the user to a certain extent
  • the target posture data used for image processing to achieve electronic stabilization is based on the imaging device at the time of image capture.
  • the motion information is determined, and therefore, the determined target pose data can reflect the pose of the image desired by the user. Therefore, in the present application, by processing the image to achieve electronic stabilization, the posture of the processed image can meet the user's expectation.
  • FIG. 1 is a schematic structural diagram of a movable platform according to an embodiment of the present application.
  • FIG. 2 is a schematic diagram of target attitude data and measurement attitude data of an electronic stabilization method according to an embodiment of the present application
  • FIG. 3 is a first schematic diagram of an image before and after being processed by an electronic stabilization method according to an embodiment of the present application
  • FIG. 4 is a second schematic diagram of an image before and after being processed by an electronic stabilization method according to an embodiment of the present application
  • FIG. 5 is a third schematic diagram of an image before and after being processed by an electronic stabilization method according to an embodiment of the present application.
  • Fig. 6 is the control principle diagram of the PTZ according to an embodiment of the present application.
  • FIG. 7 is a fourth schematic diagram of an image before and after being processed by the electronic stabilization method according to an embodiment of the present application.
  • FIG. 8 is a schematic diagram of the schematic diagram shown in FIG. 7 after reducing the size of the image B;
  • FIG. 9 is a schematic diagram of a collision limit of an electronic stabilization method according to an embodiment of the present application.
  • FIG. 10 is a schematic diagram of the influence of measurement attitude data on stabilization effect in an electronic stabilization method according to an embodiment of the present application.
  • FIG. 11 is a schematic diagram of a low-pass filter in an electronic stabilization method according to an embodiment of the present application.
  • Fig. 12 is a schematic diagram of another low-pass filtering in the electronic stabilization method according to an embodiment of the present application.
  • first and second are only used for description purposes, and cannot be interpreted as indicating or implying relative importance or implicitly indicating the number of indicated technical features.
  • features defined as “first” and “second” may expressly or implicitly include, but are not limited to, one or more of said features.
  • FIG. 1 is a schematic structural diagram of a movable platform 10 according to an embodiment of the present application, and the movable platform 10 is an example of an unmanned aerial vehicle. As shown in FIG. 1 , the image is captured by the imaging device 100 , and the imaging device 100 is carried on the movable platform 10 .
  • the imaging apparatus 100 may be a camera, a video camera, etc., specifically, a digital video camera, an action camera, a single-lens reflex camera, a mirrorless camera, etc., or may be other various aerial photography devices.
  • the imaging device may be a smart terminal with a shooting function, such as a mobile phone, a tablet, and the like with a shooting function.
  • the movable platform 10 in FIG. 1 is taken as an example of an unmanned aerial vehicle.
  • the unmanned aerial vehicle is also commonly referred to as a UAV (Unmanned Aerial Vehicle), wherein the unmanned aerial vehicle may include a fixed-wing unmanned aerial vehicle.
  • UAV Unmanned Aerial Vehicle
  • the connection between the imaging device 100 and the fuselage of the UAV is not limited to the position shown in FIG. 1 , that is, the imaging device 100 can not only be connected to the bottom of the UAV, but also can be connected to the UAV.
  • the top and side parts are connected to each other, which is not limited in this embodiment.
  • the movable platform may also be a handheld gimbal, an unmanned vehicle, a robot, or the like.
  • the handheld pan/tilt may include a handle and a pan/tilt, and the pan/tilt is used to carry the imaging device.
  • the handle can be used to support the head, and the shape of the handle is not limited to the column shape, that is, the handle can be not only cylindrical, prismatic, etc., but also truncated, pyramidal, spherical, etc., or even
  • the specific shape of the handle is not limited in this embodiment, such as the combination of the above-mentioned various shapes or the special shape.
  • Such a movable platform can be directly operated by a user by hand, has a wide range of application scenarios, is convenient for the user to operate, and saves costs.
  • the handle can be provided with an operating device, and the operating device can be an operating button, an operating lever or a control interface, etc., so as to control the imaging device or other components of the movable platform, such as controlling the opening, closing and shooting of the imaging device, etc. .
  • the unmanned vehicle may include a chassis.
  • the chassis can be used to support the imaging device, and the movement of the chassis is not limited to using wheels to move, and can also be moved through other mechanisms such as crawlers.
  • the number of wheels of the unmanned vehicle may be one or more, which is not limited in this embodiment.
  • the robot may include a body. It can be understood that the connection between the imaging device and the fuselage of the robot is not limited to the connection with the head of the fuselage of the robot, and can also be connected with other parts such as the robot arm and the back of the fuselage of the robot. This is not limited.
  • the method for electronic stabilization of an image includes: acquiring motion information of the imaging device at the current capture moment of the image and attitude data obtained based on measurement data of the attitude sensor, and the image is captured by the imaging device; at least partially based on the current motion information Determine the target attitude data for stabilizing the image; process the image according to the target attitude data and the attitude data obtained based on the measurement data of the attitude sensor at the current capture moment.
  • the attitude data obtained based on the measurement data of the attitude sensor at the current capture moment is directly obtained or obtained by fusion of the measurement data of the attitude sensor, and the attitude sensor includes but is not limited to an inertial measurement unit.
  • the motion information of the imaging device can include one of moving angular velocity and moving angular acceleration, and can also be obtained by various sensors, such as gyroscopes or accelerometers, and can also be obtained by estimating the angular velocity information of the images of the frames before and after the capture moment, This embodiment does not limit this.
  • the motion information may also include one of translational speed and translational acceleration, and the translational speed or translational acceleration may be obtained through a linear motion sensor, such as a global positioning system GPS.
  • the motion information also includes one of translation speed and translation acceleration
  • the position data of the imaging device at the current capture moment of the image can also be obtained, so as to further process the image in combination with the position data at the current capture moment, so as to achieve electronic stabilization.
  • the moving angular velocity and the moving angular velocity mainly consider the influence of the rotational motion of the imaging device on the imaging quality
  • the translational speed and translational acceleration mainly consider the influence of the translational motion of the imaging device on the imaging quality.
  • the following contents of the embodiments of the present application are exemplified by the influence of the rotational motion of the imaging device on the imaging quality.
  • FIG. 2 is a schematic diagram of target attitude data b (dotted line in the figure) and attitude data a (solid line in the figure) obtained based on the measurement data of the attitude sensor of the electronic stabilization method according to an embodiment of the present application, as shown in FIG. 2 where the abscissa t represents the time, and the ordinate v represents the attitude.
  • the target attitude data is sine in the coordinate system shown in FIG. 2 as an example, but it can be understood that in other embodiments, the target The attitude data in the coordinate system shown in FIG. 2 can be in various other shapes, which can be related to the user’s control of the movable platform and/or the PTZ, and can be attitude data obtained based on the measurement data of the attitude sensor obtained by smoothing.
  • Fig. 2 takes the situation that the target attitude data b and the attitude data a obtained based on the measurement data of the attitude sensor are only different at the time t1, and are consistent at other times as an example.
  • the imaging device When the imaging device is carried on the movable platform and captures images, the imaging device may be affected by various factors such as wind force, air pressure changes, etc. due to the movable platform or the pan/tilt, resulting in undesired movements of the user, such as shaking,
  • the posture of the imaging device when capturing the image is not expected by the user.
  • the target posture data and the posture data obtained based on the measurement data of the posture sensor are consistent, and at this time, the posture when the imaging device captures the image is expected by the user.
  • the imaging device can also be used independently without being carried on the movable platform. In this case, when the imaging device is operated by the user by hand, the user may also shake the user's hand. sports.
  • processing the image according to the target posture data and the posture data at the current capture moment may be to calculate the posture compensation amount at the moment according to the target posture data and the posture data at the current capture moment. Then, the image before the electronic stabilization processing is corrected according to the attitude compensation amount to obtain the processed image.
  • the position compensation amount at this moment can also be calculated according to the target position data and the position data at the current capture moment, so as to adjust the image before electronic stabilization processing according to the position compensation amount Corrections are made to obtain the processed image.
  • the above correction may be implemented by cropping the image before the electronic stabilization processing.
  • the above correction may include rotation correction and/or translation correction.
  • the above-mentioned correction when considering the influence of the rotational motion of the imaging device on the imaging quality, the above-mentioned correction may include rotational correction and/or translational correction; when considering the influence of the translational motion of the imaging device on the imaging quality, the above-mentioned correction may include translational correction. Correction.
  • FIG. 3 is a first schematic diagram of an image before and after being processed by an electronic stabilization method according to an embodiment of the present application.
  • image A represents the image before processing by the electronic stabilization method
  • image B represents the image before processing by the electronic stabilization method.
  • the preset position may be located in the middle of the image A, thereby ensuring that when the target attitude data and the attitude data obtained based on the measurement data of the attitude sensor are inconsistent, the image A can be in various directions. All have suitable clipping space. In other embodiments, the preset position may not be located in the middle of the image A.
  • the cropping of the image before the electronic stabilization process according to the attitude compensation amount may be: the position of the cropped image in the pre-cropped image deviates from the direction of the preset position, and the measurement based on the attitude sensor at the corresponding moment is different.
  • the pose obtained from the data deviates from the target pose in the opposite direction.
  • FIG. 4 is a second schematic diagram of an image before and after being processed by the electronic stabilization method according to an embodiment of the present application.
  • the direction in which the posture obtained based on the measurement data of the posture sensor at the current capture moment of the image A corresponding to FIG. 4 deviates from the target posture is upward, then the direction in which the position of the cropped image in the pre-cropped image deviates from the preset position is to Down.
  • the position of the cropped image in the pre-cropped image deviates from the direction of the preset position. for upward.
  • FIG. 5 is a third schematic diagram of an image before and after being processed by the electronic stabilization method according to an embodiment of the present application.
  • the direction in which the measurement posture of the image A corresponding to FIG. 5 at the current capture moment deviates from the target posture is counterclockwise rotation, and the direction in which the position of the cropped image in the pre-cropped image deviates from the preset position is clockwise rotation.
  • the direction in which the measurement posture of the image A at the current capture moment deviates from the target posture is clockwise rotation
  • the direction in which the position of the cropped image in the pre-cropped image deviates from the preset position is counterclockwise rotation.
  • the image when the target posture data corresponds to the movement of the imaging device expected by the user, the image can be processed according to the target posture data to achieve electronic stabilization, and an image with the posture expected by the user can be obtained.
  • the motion information of the imaging device at the time of capturing the image can reflect the posture of the image expected by the user to a certain extent
  • the target posture data for processing the image to achieve electronic stabilization is based on the imaging device at the time of image capturing.
  • the motion information is determined, and therefore, the determined target pose data can reflect the pose of the image desired by the user. Therefore, in the present application, by processing the image to achieve electronic stabilization processing, the posture of the processed image can meet the user's expectation.
  • determining the target attitude data for stabilizing the image according to the moving angular velocity may include: determining the target attitude data for stabilizing the image according to a comparison result of the moving angular velocity and an angular velocity threshold.
  • the target attitude data used to stabilize the image can be It avoids various complicated situations caused by various specific values of the moving angular velocity, and makes the process of determining the target attitude data simple.
  • the following describes the influence of the moving angular velocity of the imaging device on image stabilization and the processing strategy:
  • the electronic stabilization method may further include: acquiring an attitude control instruction, where the attitude control instruction is used to control the target attitude of the imaging device. Then, according to the comparison result between the moving angular velocity and the angular velocity threshold, determining the target attitude data for stabilizing the image may include: when the moving angular velocity is less than or equal to the angular velocity threshold (at this time, the movable platform can be in a hovering state or a low-speed motion state) , and determine that the target posture data for stabilizing the image is the target posture of the imaging device corresponding to the posture control instruction.
  • the target posture of the imaging device corresponding to the posture control instruction can represent the target posture of the imaging device expected by the user. Therefore, by determining the target posture data for stabilizing the image as the target posture of the imaging device corresponding to the posture control command , an image with the user's desired posture can be obtained, that is, after processing by the electronic stabilization method, the influence of the user's undesired motion on the image posture during the shooting of all images can be eliminated, and the user experience can be improved.
  • This method of determining the target attitude data for stabilizing the image compared with the method of processing the measured attitude through a filter to obtain the target attitude data for stabilizing the image, can be equivalent to realizing the noise analysis of different frequency bands. Elimination, even in the case of relatively high requirements on the picture, such as the case of high-magnification zoom, the image after the electronic stabilization process will not be perceptible to the human eye.
  • the gesture control instruction can be issued by the user in various ways, for example, through a control terminal of the movable platform, such as a remote control, various smart devices, and the like.
  • a control terminal of the movable platform such as a remote control, various smart devices, and the like.
  • the gesture control command is issued by the user through the remote controller of the movable platform, it can be issued through buttons, joysticks, touch screens, etc. on the remote controller.
  • the gesture control instruction is issued by the user through the smart device, it can be issued through the APP on the smart device, wherein the smart device can be a mobile phone, a tablet, a computer, or the like.
  • the posture control instructions are used to instruct the control of the target posture of the imaging device to be achieved by controlling the posture of the movable platform.
  • the posture control instruction is used to instruct the control of the target posture of the imaging device by controlling the posture of the pan-tilt (as shown in FIG. 1 , the movable platform may include the pan-tilt 200 ), and the imaging device passes through the pan-tilt. carried on a movable platform.
  • the pan/tilt may include one pan/tilt part, two pan/tilt components, three pan/tilt components or more pan/tilt components, and accordingly, the pan/tilt may allow the imaging device to wrap around one, two, three or more pan/tilt parts axis rotation, the axes used for rotation may or may not be orthogonal to each other.
  • the pan/tilt component can control the posture of the imaging device through the motor, including controlling one or more of the pitch angle, roll angle, and yaw angle of the imaging device.
  • the imaging device can rotate around the pitch axis, Rotation of one or more of the roll and yaw axes.
  • each pan-tilt part may include a connecting arm.
  • the first pan-tilt part is connected to the body of the movable platform, and the first pan-tilt part can be rotated relative to the body, so that the yaw angle of the imaging device changes, that is, when the first connecting arm rotates relative to the body , the imaging device can be rotated around the yaw axis.
  • the second pan-tilt part is connected to the first pan-tilt part, and the second pan-tilt part can be rotated relative to the main body, so that the roll angle of the imaging device changes, that is, when the second pan-tilt part is rotated relative to the main body, the imaging device can be rotated Rotate around the roll axis.
  • the third pan-tilt part is connected with the second pan-tilt part, and the third pan-tilt part can be rotated relative to the main body, so that the pitch angle of the imaging device changes, that is, when the third pan-tilt part is rotated relative to the main body, the imaging device can be rotated around the main body. Pitch axis rotation.
  • the pan/tilt head may only include one pan/tilt member, and the pan/tilt member may be rotated relative to the body to change the yaw angle of the imaging device, that is, when the pan/tilt member is rotated relative to the body, imaging The device rotates around the yaw axis.
  • the gimbal connected to the body can have one gimbal part, two gimbal parts, three gimbal parts or more gimbal parts, and can make imaging.
  • the device is rotated about one, two, or three of the pitch, roll, and yaw axes, so that the imaging device can also be rotated about more axes, and the like.
  • the gimbal can also have one gimbal part, two gimbal parts, three or more gimbal parts, and the imaging device can be rotated around the pitch axis One or two or three of the , the roll axis and the yaw axis are rotated so that the imaging device can also be rotated about more than three axes, etc. That is to say, regardless of the type of movable platform, the gimbal can be a single-axis gimbal, a dual-axis gimbal, a three-axis gimbal or a gimbal with other axes.
  • FIG. 6 is a control principle diagram of a pan-tilt head according to an embodiment of the present application.
  • the gimbal adjusts the posture of the imaging device, it detects the current posture of the imaging device through the inertial measurement element, and compares the current posture of the imaging device with the target posture to obtain the control deviation.
  • the control deviation controls the motor to adjust the posture of the imaging device, and finally reduces the control deviation to ensure that the deviation between the actual posture of the imaging device and the target posture is as small as possible.
  • the gimbal for realizing mechanical stabilization can be combined with the electronic stabilization method for processing images, so as to avoid the gimbal's control accuracy and the measurement accuracy of the inertial measurement element of the gimbal.
  • the stabilization effect is not stable and other problems occur.
  • the cropping of the image before the electronic stabilization process according to the attitude compensation amount may be: the degree to which the position of the cropped image in the pre-cropped image deviates from the preset position, and the measured attitude at the corresponding moment deviates from the target.
  • the degree of attitude is proportional. That is, the greater the degree to which the measured posture deviates from the target posture, the greater the degree to which the position of the cropped image in the pre-cropped image deviates from the preset position. Thereby, it can be guaranteed that the processed image is what the user expects.
  • FIG. 7 is a fourth schematic diagram of an image before and after being processed by the electronic stabilization method according to an embodiment of the present application.
  • the direction in which the measured posture at the capture moment of the image A corresponding to FIG. 7 deviates from the target posture is counterclockwise rotation, and the direction in which the position of the cropped image in the pre-cropped image deviates from the preset position is clockwise rotation.
  • the area C shown in FIG. 7 may appear, while the image A does not have the area C. Therefore, at this time, the processed image B will appear undesired by the user. area C.
  • the measurement attitude will deviate from the target attitude due to the lag of the movable platform or the gimbal in response to the attitude control command output by the user. If the degree is large, it is easy to cause the above-mentioned area C to appear in the processed image B.
  • determining the target attitude data for stabilizing the image may include: when the moving angular velocity is greater than the angular velocity threshold (at this time, the movable platform can be in a large maneuvering motion) , and according to the attitude data obtained based on the measurement data of the attitude sensor at the current capture moment and the preset bias threshold, the target attitude data for stabilizing the image is determined.
  • the processed image B is prevented from appearing in the area C that the user does not expect, so as to improve the user experience.
  • the preset bias threshold is used to ensure the continuity of content changes in the images captured by the imaging device, so as to reduce the influence of external disturbances on the imaging quality of the imaging device.
  • the angular velocity threshold may be related to the difference in size of the images before and after processing. Since the angular velocity threshold is related to avoiding the user-undesired area C in the processed image B, and in the case that the measured attitude deviates from the target attitude to the same degree, whether the processed image B has the user-undesired area C is related to the image being processed.
  • the size difference before and after is related. For example, please refer to FIG. 7 and FIG. 8 at the same time.
  • FIG. 8 is a schematic diagram of the schematic diagram shown in FIG. 7 after the size of the image B is reduced.
  • the processed image B may appear in the area C that the user does not expect (as shown in Figure 7), or may not appear in the area that the user does not expect.
  • C (as shown in Fig. 8)
  • Fig. 7 and Fig. 8 it can be seen from Fig. 7 and Fig. 8 that if the attitude obtained based on the measurement data of the attitude sensor deviates from the target attitude to the same degree, whether there is an area C that is not expected by the user in the processed image B and Image size differences before and after processing are correlated. Therefore, by correlating the angular velocity threshold with the size difference of the images before and after processing, the above-mentioned region C can be less likely to appear in the processed image B.
  • the angular velocity threshold can be positively correlated with the size difference of the images before and after processing. That is to say, the larger the size difference between the images before and after processing, the larger the angular velocity threshold may be, and the smaller the size difference between the images before and after processing, the smaller the angular velocity threshold may be.
  • the size difference of the images before and after processing is larger, the above-mentioned area C will appear only when the attitude attitude obtained based on the measurement data of the attitude sensor deviates from the target attitude.
  • the target posture data used to stabilize the image can be made to be the target posture of the imaging device corresponding to the posture control instruction to the greatest extent possible.
  • the target attitude data used to stabilize the image is the target attitude of the imaging device corresponding to the attitude control command
  • the noise in different frequency bands can be eliminated, even in the case of high requirements on the picture, such as high magnification zoom Under these conditions, the image after the electronic stabilization process will not be perceptible to the human eye.
  • the electronic stabilization method of this embodiment may further include: acquiring attitude data obtained based on the measurement data of the attitude sensor of the imaging device at multiple different times; then determining the target attitude for the stabilization image according to the moving angular velocity
  • the data may include: according to the moving angular velocity, screening the attitude data obtained based on the measurement data of the attitude sensors of the imaging devices at different times; and determining the target attitude data for stabilizing the image according to the screened attitude data.
  • the target attitude data for stabilizing the image when determining the target attitude data for stabilizing the image, the attitude data obtained based on the measurement data of the attitude sensor of the imaging device at different times are used, and the change of the imaging device that the user does not expect is usually only in the A small part of the time period is generated, that is to say, the attitude data obtained based on the measurement data of the attitude sensor of the imaging device at multiple different times generally corresponds to the attitude of the imaging device expected by the user. Therefore, by determining the target posture data for image stabilization at the image capturing moment through the attitude data obtained based on the measurement data of the attitude sensor of the imaging devices at different times, the target for image stabilization can be made to a certain extent. Attitude data meets user expectations.
  • screening the attitude data obtained based on the measurement data of the attitude sensor of the imaging device at different times may include: determining the cutoff frequency of low-pass filtering according to the moving angular velocity; The attitude data obtained based on the measurement data of the attitude sensor of the imaging device at the moment is subjected to low-pass filtering to obtain the filtered attitude data.
  • Low-pass filtering is a filtering method. The rule is that low-frequency signals can pass normally, while high-frequency signals that exceed the set critical value (cutoff frequency) are blocked and weakened. But the magnitude of blocking, attenuation can be changed according to different frequencies and different filtering procedures (purposes). It will be appreciated that low-pass filtering is sometimes referred to as high-frequency removal filtering or maximum removal filtering.
  • Using the filtering method to process the attitude data obtained based on the measurement data of the attitude sensor of the imaging devices at different times can effectively eliminate the influence of the user's undesired motion (for example, shaking, etc.) on the image attitude during the image shooting process. .
  • determining the cut-off frequency of the low-pass filtering according to the moving angular velocity may include: determining the cut-off frequency of the low-pass filtering according to the comparison result between the moving angular velocity and the moving angular velocity threshold.
  • the moving angular velocity of the movable platform 10 can be obtained through various sensors, and can also be obtained through estimation of angular angular velocity information of images of frames before and after the capture moment, which is not limited in this embodiment.
  • the moving angular velocity of the movable platform 10 can be obtained through a command input by the user for controlling the moving angular velocity of the imaging device, for example, when the user inputs the command for controlling the moving angular velocity of the imaging device through a joystick , the moving angular velocity of the movable platform 10 can be obtained by the rod amount of the rocker.
  • determining the cut-off frequency of the low-pass filtering may include: when the moving angular velocity is less than or equal to the moving angular velocity threshold, determining the cut-off frequency of the low-pass filtering as a preset cut-off frequency.
  • determining the cut-off frequency of the low-pass filtering may include: when the moving angular velocity is greater than the moving angular velocity threshold, adjusting the cut-off frequency of the low-pass filtering to be greater than a preset cut-off frequency.
  • the area C shown in FIG. 7 may appear, but the image A does not have area C.
  • Image B appears in an area C that is not intended by the user.
  • the moving angular velocity is greater than the moving angular velocity threshold, the attitude obtained based on the measurement data of the attitude sensor will deviate from the target attitude to a greater extent due to the hysteresis of the movable platform or the gimbal in responding to the attitude control commands output by the user, and it is easy to The above-mentioned area C is made to appear in the processed image B.
  • the target posture can be properly corrected after the target posture is determined by the above method, so that the above-mentioned area C will not appear when the image is processed according to the corrected target posture. Since the above-mentioned area C appears after the image is rotated clockwise by an angle corresponding to the target posture before correction, when the image is rotated clockwise, the rotation angle can correspond to the corrected target posture. It can be understood that the angle corresponding to the target posture before correction is greater than the angle corresponding to the target posture after correction, so that the image B can be prevented from appearing in the above-mentioned area C (as shown in FIG. 9 ).
  • Fig. 9 is an implementation according to the present application.
  • the cut-off frequency of the low-pass filtering is lower, the inconsistency between the measurement attitude and the target attitude is more serious, especially when the moving angular velocity of the imaging device is faster, the attitude obtained based on the measurement data of the attitude sensor deviates from the target.
  • the cut-off frequency is greater than the preset cut-off frequency, so as to avoid the above-mentioned bumping into the limit or prevent the processed image B from appearing in the above-mentioned area C, so as to improve the user experience.
  • the difference between the cut-off frequency of the low-pass filter and the preset cut-off frequency may be greater. Because, when the moving angular velocity of the imaging device is faster, the more serious the situation that the attitude obtained based on the measurement data of the attitude sensor deviates from the target attitude is, the more likely the above phenomenon of hitting the limit will occur or the more likely the processed The above-mentioned area C appears in the image B.
  • the difference between the cut-off frequency of the low-pass filter and the preset cut-off frequency is greater, so as to avoid the phenomenon of hitting the limit or avoid
  • the above-mentioned area C appears in the processed image B, so as to improve the user experience.
  • the above-mentioned moving angular velocity threshold and angular velocity threshold may be different or the same.
  • the above-mentioned corresponding embodiments of determining the target attitude data for stabilizing the image according to the moving angular velocity may be coupled.
  • Corresponding embodiments can be selected and used as required.
  • this method can use attitude control instructions to eliminate noise in different frequency bands; if the imaging device is applied independently or directly carried on a movable platform, it can be based on the moving angular velocity and the moving angular velocity threshold. By comparison, the target attitude data for stabilizing the image is determined, so that the image stabilization can still be better achieved without the gimbal bearing.
  • the target posture of the imaging device can also be realized by the posture control of the movable platform, but in this application scenario, the movable platform and the imaging device are fixedly connected, and the movable platform What is more realized is the change of the trajectory, not mainly used for the attitude adjustment of the imaging device, while the gimbal is mainly used for the attitude adjustment of the imaging device.
  • the target attitude data is used, the situation where the imaging device is mounted on the movable platform via the gimbal can be mainly considered.
  • the moving angular velocity obtained by integrating the moving angular acceleration may be used for comparison with the angular velocity threshold and the moving angular velocity threshold.
  • the moving angular acceleration may be directly compared with a set threshold value. The specific judgment method and judgment result are similar to those described above, and will not be repeated here.
  • the movable platform may be provided with a filter with a variable cut-off frequency, so as to adjust the cut-off frequency of the low-pass filter by adjusting the cut-off frequency of the filter.
  • the movable platform may be provided with a plurality of filters with different cutoff frequencies. During low-pass filtering, the filter corresponding to the moving angular velocity is selected to perform low-pass filtering, or the low-pass filtering results of different filters are performed. Weighted average to obtain the target attitude data corresponding to the cutoff frequency determined according to the moving angular velocity.
  • the attitude data obtained based on the measurement data of the attitude sensor of the imaging devices at different times may include: attitude data at the current capture time.
  • the measurement attitude data of the imaging device at the current capture moment of the image can reflect the attitude of the imaging device expected by the user at this moment to a certain extent, so that the attitude data obtained based on the measurement data of the attitude sensor of the imaging devices at different times includes this data.
  • the obtained target pose data can be made to meet user expectations.
  • At least part of the plurality of different times is located before the current capture time, and at least another part of the plurality of different times is located after the current capture time.
  • the duration corresponding to the at least part of the multiple different moments may be equal to the duration corresponding to the at least another part of the multiple different moments.
  • FIG. 10 is a schematic diagram of the influence of attitude data obtained based on the measurement data of the attitude sensor on the stabilization effect in the electronic stabilization method according to an embodiment of the present application.
  • the attitude obtained based on the measurement data of the attitude sensor is shown in Fig. 10(a), in which the actual exposure point ET in Fig.
  • the capture time T0, the attitude data obtained based on the measurement data of the attitude sensor includes the captured The first attitude data in the first time period T1 before the time T0, and the second attitude data in the second time period T2 after the time T0 are captured.
  • the time used for collecting the first attitude data is Delta-t1
  • the first attitude data corresponds to the actual attitude IT
  • the target attitude IE can be obtained based on the first attitude data.
  • the frequency domain low-pass filter is used as the median filter
  • the median point EE of the first attitude data is located at T1/2. Since the image at the actual exposure point ET is subsequently stabilized by the data at the median point EE, there is a delay Delta-t2 between the median point EE and the actual exposure point ET, where Delta-t2 is equal to T1/2.
  • the delay is due to the processing result obtained by the low-pass filter, that is, the deviation between the median point EE and the actual exposure point ET, or it can be understood as a filtering error. If the imaging device jitters during the delay Delta-t2, the median value is used. The data at point EE will deviate from the stabilization of the image to be stabilized.
  • the time used for the collection of the first attitude data and the second attitude data is Delta-t1
  • the first attitude data and the second attitude data correspond to the actual attitude IT
  • T1 and T2 are equal as an example
  • the target posture IE can be obtained based on the first posture data and the second posture data.
  • the frequency domain low-pass filter is used as the median filter
  • the median point EE of the first attitude data and the second attitude data is located at T0, that is, it coincides with the actual exposure point ET, so that (b) can be avoided.
  • the delay Delta-t2 of T1/2 exists in the figure, that is, the delay Delta-t2 is equal to 0. In this way, it is more accurate to use the data at the median point EE to stabilize the image.
  • At least part of the multiple different moments is located before the current capture moment, at least another part of the multiple different moments is located after the current capture moment, and the duration corresponding to the at least part of the multiple different moments is different from that of the multiple different moments.
  • the duration corresponding to the at least another part of the time can be set to be equal to ensure the stabilization effect on the image.
  • the influence of the delay of the low-pass filtering on the filtering result caused by the collection of the first attitude data and the second attitude data can be compensated, so that the low-pass filtering can have a lower cut-off frequency to process more data, Understandably, when the cutoff frequency is lower, the picture of the image is smoother.
  • the number of attitude data obtained based on the measurement data of the attitude sensor of the imaging device at different times may be greater than 2, so as to improve the effect of stabilization. It can be understood that the attitude data obtained based on the measurement data of the attitude sensor of the imaging device at multiple different times are used in determining the target attitude data for stabilizing the image, and the changes of the imaging device that are not expected by the user are usually only one time. A small part of the time period is generated, that is to say, the attitude data obtained based on the measurement data of the attitude sensor of the imaging device at multiple different times usually corresponds to the attitude of the imaging device expected by the user.
  • Determining the target attitude data used for stabilizing the image at the current capture moment of the imaging device based on the attitude data obtained from the measurement data of the attitude sensor of the imaging device at the moment can make the target attitude data used for stabilizing the image meet user expectations to a certain extent, The more attitude data obtained based on the measurement data of the attitude sensor of the imaging device at different times, the more the influence of the undesired motion at the capture time on the result can be eliminated. The more attitude data obtained from the measurement data, the smoother the picture of the processed image and the better the effect of electronic stabilization.
  • Performing low-pass filtering on the attitude data obtained based on the measurement data of the attitude sensor of the imaging apparatus at different times according to the cutoff frequency of the low-pass filtering may include: obtaining the measurement data based on the attitude sensor of the imaging apparatus at different times The pose data is subjected to multiple sub-low pass filtering.
  • the calculation process of the low-pass filtering can be performed separately, so as to avoid the phenomenon of wasting computing resources due to the centralized calculation, and relieve the calculation pressure.
  • all of the attitude data obtained from the measurement data based on the attitude sensor of the imaging devices at different times are still used, so it is possible to achieve the same measurement results as the measurement data based on the attitude sensors of the imaging devices at different times.
  • the pose data obtained from the data is filtered once to the same or similar effect.
  • the last sub-low-pass filtering (may be the sub-low-pass filtering in which the output data is related to the measurement attitude data at the multiple different times, that is, the output data of the last sub-low-pass filtering is the measurement attitude data at the multiple different times)
  • the cutoff frequency is determined according to the cutoff frequency of the low-pass filtering.
  • the cut-off frequency of any sub-low-pass filtering except the cut-off frequency of the last sub-low-pass filtering is greater than the cut-off frequency of the low-pass filtering.
  • any sub-low-pass filtering After any sub-low-pass filtering outputs corresponding output data, the input data of any sub-low-pass filtering is deleted. In this way, the problem that data occupies a large storage space can be avoided, and the storage pressure can be alleviated.
  • the input data of any sub-low-pass filter After any sub-low-pass filter outputs the corresponding output data, the input data of any sub-low-pass filter is deleted, that is, in the corresponding calculation process.
  • the input data of any sub-low-pass filtering is stored, so that the corresponding input data can be obtained immediately when the corresponding input data needs to be obtained again when an accident occurs.
  • the input data of any sub-low-pass filtering may be deleted during the process of outputting the corresponding output data by the sub-low-pass filtering, or when the sub-low-pass filtering is being calculated.
  • FIG. 11 is a schematic diagram of a low-pass filter in an electronic stabilization method according to an embodiment of the present application, wherein d1, d2, d3, d4, d5, d6, d7 and d8 respectively correspond to an imaging device at a moment
  • the attitude data obtained based on the measurement data of the attitude sensor, f1, f2, f3, f4, f5, f6 and f7 respectively represent different sub-low-pass filtering. As shown in FIG.
  • At least part of the input data of the sub-low-pass filtering is the attitude data ( d1, d2, d3, d4, d5, d6, d7, and d8 correspond to the part of the attitude data obtained based on the measurement data of the attitude sensor of the imaging device), and at least another part of the sub-low-pass filtering (sub-elements represented by f5, f6 and f7)
  • the input data of low-pass filtering is the output data of other sub-low-pass filtering.
  • the sub-low-pass filtering of attitude data obtained based on the measurement data of the attitude sensor of the imaging device as input data at different times can be called first-order filtering (sub-low-pass filtering represented by f1, f2, f3 and f4), and the The sub-low-pass filtering in which the input data is the output data of the first-stage filtering is called the second-stage filtering (sub-low-pass filtering represented by f5 and f6), and the sub-low-pass filtering in which the input data is the output data of the second-stage filtering is called the third-stage filtering.
  • stage filtering sub-low pass filtering denoted by f7, and so on.
  • the input data of all sub-low-pass filtering in the at least another part of the sub-low-pass filtering is the output data of the at least part of the sub-low-pass filtering.
  • FIG. 12 is a schematic diagram of another low-pass filtering in the electronic stabilization method according to an embodiment of the present application. That is, the low-pass filtering only includes the first-stage filtering (sub-low-pass filtering represented by f1, f2, f3, and f4) and the second-stage filtering (sub-low-pass filtering represented by f5), thereby simplifying the low-pass filtering process.
  • Performing multiple sub-low-pass filtering on the attitude data obtained based on the measurement data of the attitude sensors of the imaging devices at different times may include: every time a preset number of the imaging devices are acquired based on the attitude sensor measurement data, the attitude obtained data, the corresponding sub-low-pass filtering is performed. For example, for every 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, When 17, 18, 19, 20, 21, 22, 23, 24, 25, 26 or more imaging devices obtain attitude data based on the measurement data of attitude sensors, corresponding The sub-low-pass filtering of , thus, facilitates the calculation of the secondary filter.
  • Performing multiple sub-low-pass filtering on the attitude data obtained based on the measurement data of the attitude sensor of the imaging apparatuses at different times may further include: obtaining the attitude data obtained from the measurement data of the imaging apparatus based on the attitude sensor corresponding to the acquired preset time period. When the attitude data is used, the corresponding sub-low-pass filtering is performed.
  • the preset duration can be 0.1s, 0.2s, 0.3s, 0.4s, 0.5s, 0.6s, 0.7s, 0.8s, 0.9s, 1.0s, 1.1s, 1.2s, 1.3s, 1.4s, 1.5s, 1.6s, 1.7s, 1.8s, 1.9s, 2.0s, 2.1s, 2.2s, 2.3s, 2.4s, 2.5s, etc.
  • the total duration corresponding to the attitude data obtained based on the measurement data of the attitude sensors of the imaging devices at different times is 2s, and the low-pass filtering only includes the first-level filtering and the second-level filtering as an example. illustrate.
  • the specific process of low-pass filtering may include: using the attitude data obtained based on the measurement data of the attitude sensor within the first 0.1s to perform the first first-level filtering, and deleting the first-level filtering after the output result of the first-level filtering. filter the input data.
  • the eighth first-stage filtering is performed using the attitude data obtained based on the measurement data of the attitude sensor within the eighth 0.1s, and the input data of the eighth first-stage filtering is deleted after the output result of the eighth first-stage filtering.
  • the ninth first-stage filtering is performed using the attitude data obtained based on the measurement data of the attitude sensor within the ninth 0.1s, and the input data of the ninth first-stage filtering is deleted after the output result of the ninth first-stage filtering.
  • the tenth first-stage filtering is performed using the attitude data obtained based on the measurement data of the attitude sensor within the tenth 0.1s, and the input data of the tenth first-stage filtering is deleted after the output result of the tenth first-stage filtering.
  • the total duration corresponding to the gesture data obtained based on the measurement data of the gesture sensors of the imaging devices at different times may be longer than the duration corresponding to the image (the image may be a video), for example, the duration corresponding to the image 2 times, 3 times, 4 times, 5 times, etc., so as to ensure the stabilization effect of the image.
  • the imaging device can be carried on the movable platform through the pan/tilt, and the pan/tilt is used to stabilize the imaging device.
  • a pan/tilt for realizing mechanical stabilization can be combined with an electronic stabilization method for processing images, so as to prevent the pan/tilt from being caused by the control accuracy of the pan/tilt and the measurement accuracy of the inertial measurement element of the pan/tilt.
  • the stabilization effect is not stable and other problems occur.
  • This embodiment also provides a computer-readable storage medium, where the computer-readable storage medium stores executable instructions, and when executed by one or more processors, the executable instructions can cause one or more processors to execute any of the above An electronic stabilization method.
  • the computer-readable storage medium may also be referred to as a memory, and the executable instructions may also be referred to as a program.
  • the processor may perform various appropriate actions and processes according to programs stored in read only memory (ROM) or loaded into random access memory (RAM).
  • a processor may include, for example, a general-purpose microprocessor (eg, a CPU), an instruction set processor and/or a related chipset, and/or a special-purpose microprocessor (eg, an application specific integrated circuit (ASIC)), among others.
  • the processor may also include onboard memory for caching purposes.
  • the processor may comprise a single processing unit or multiple processing units for performing different actions of the method flow according to this embodiment.
  • the processor, ROM, and RAM are connected to each other through a bus.
  • the processor performs various operations of the method flow according to the present embodiment by executing programs in the ROM and/or RAM. Note that programs may also be stored in one or more memories other than ROM and RAM.
  • the processor may also perform various operations of the method flow according to the present embodiment by executing programs stored in one or more memories.
  • the apparatus to which the computer-readable storage medium is applied may further include an input/output (I/O) interface, which is also connected to the bus.
  • the device employing the computer-readable storage medium may also include one or more of the following components connected to the I/O interface: an input portion including a keyboard, a mouse, etc.; an input portion such as a cathode ray tube (CRT), a liquid crystal display (LCD) ), etc., and an output section for speakers, etc.; a storage section including a hard disk, etc.; and a communication section including a network interface card such as a LAN card, a modem, and the like.
  • the communication section performs communication processing via a network such as the Internet.
  • Removable media such as magnetic disks, optical disks, magneto-optical disks, semiconductor memories, etc., are mounted on the drive as needed, so that the computer program read therefrom is installed into the storage section as needed.
  • the method flow according to this embodiment can be implemented as a computer software program.
  • the present embodiment includes a computer program product comprising a computer program carried on a computer-readable storage medium, the computer program containing program code for performing the method shown in the flowchart.
  • the computer program may be downloaded and installed from a network via the communication portion, and/or installed from a removable medium.
  • the above-described functions defined in the system of the present embodiment are executed.
  • computer readable storage media may include, but are not limited to, non-volatile or volatile storage media such as random access memory (RAM), static RAM, dynamic RAM, read only memory (ROM), programmable ROM , Erasable Programmable ROM, Electrically Erasable Programmable ROM, Flash Memory, Secure Digital (SD) Card, etc.
  • RAM random access memory
  • ROM read only memory
  • programmable ROM Erasable Programmable ROM
  • Flash Memory Flash Memory
  • SD Secure Digital
  • This embodiment also provides an electronic stabilization device, which includes a memory and a processor.
  • Memory is used to store executable instructions.
  • the processor is configured to execute executable instructions stored in the memory to perform the following operations: acquiring motion information of the imaging device at the current capture moment of the image and attitude data obtained based on measurement data of the attitude sensor, the image is captured by the imaging device; at least part of the image is captured by the imaging device;
  • the target attitude data for stabilizing the image is determined according to the motion information; the image is stabilized according to the target attitude data and the attitude data obtained based on the measurement data of the attitude sensor at the current capture moment.
  • the electronic stabilization device may be located on the movable platform, or may be independent of the movable platform and be communicatively connected with the movable platform.
  • the electronic stabilization device can also be an independent electronic device without any association with the movable platform, such as integrated in the imaging device.
  • the motion information includes the moving angular velocity
  • the processor may specifically perform the following operations: according to the comparison result between the moving angular velocity and the angular velocity threshold, determine the target attitude data for stabilizing the image.
  • the processor may also perform the following operations: obtain an attitude control instruction, which is used to control the target attitude of the imaging device; when the moving angular velocity is less than or equal to the angular velocity threshold, determine the target attitude data used for stabilizing the image to be the same as the attitude control instruction.
  • the target pose of the corresponding imaging device may also perform the following operations: obtain an attitude control instruction, which is used to control the target attitude of the imaging device; when the moving angular velocity is less than or equal to the angular velocity threshold, determine the target attitude data used for stabilizing the image to be the same as the attitude control instruction.
  • the target pose of the corresponding imaging device may also perform the following operations: obtain an attitude control instruction, which is used to control the target attitude of the imaging device; when the moving angular velocity is less than or equal to the angular velocity threshold, determine the target attitude data used for stabilizing the image to be the same as the attitude control instruction.
  • the target pose of the corresponding imaging device may also perform the following operations: obtain an attitude control instruction, which is used to control
  • the attitude control instruction can be used to instruct to realize the control of the target attitude of the imaging device by controlling the attitude of the movable platform, the imaging device is carried on the movable platform, and/or, the attitude control instruction can be used to instruct to realize the realization by controlling the attitude of the PTZ
  • the imaging device is carried on the movable platform through the PTZ.
  • the processor may perform the following operations: when the moving angular velocity is greater than the angular velocity threshold, determine the target attitude data for stabilizing the image according to the measured attitude data of the imaging device at the time of capturing the image and the predetermined offset threshold.
  • the angular velocity threshold can be related to the difference in size of the image before and after processing.
  • the angular velocity threshold can be positively correlated with the size difference of the image before and after processing.
  • the motion information includes the moving angular velocity
  • the processor may specifically perform the following operations: acquiring attitude data obtained from the measurement data of the imaging device at different times based on the attitude sensor; The attitude data obtained from the measured data is screened; according to the screened attitude data, the target attitude data for stabilizing the image is determined.
  • the processor may further perform the following operations: determining the cutoff frequency of the low-pass filtering according to the moving angular velocity; performing low-pass filtering on the attitude data obtained based on the measurement data of the attitude sensor of the imaging device at different times according to the cut-off frequency of the low-pass filtering, to get the filtered attitude data.
  • the processor may perform the following operations: determine the cutoff frequency of the low-pass filtering according to the comparison result between the moving angular velocity and the moving angular velocity threshold.
  • the processor may perform the following operations: when the moving angular velocity is less than or equal to the moving angular velocity threshold, determine that the cutoff frequency of the low-pass filtering is a preset cutoff frequency.
  • the processor may perform the following operations: when the moving angular velocity is greater than the moving angular velocity threshold, adjust the cutoff frequency of the low-pass filtering to be greater than the preset cutoff frequency.
  • the difference between the cut-off frequency of the low-pass filter and the preset cut-off frequency can be greater.
  • the attitude data obtained based on the measurement data of the attitude sensor of the imaging devices at different times may include: attitude data obtained based on the measurement data of the attitude sensor at the current capture moment.
  • At least a portion of the plurality of different moments is located before the current capture moment; at least another portion of the plurality of different moments is located after the current capture moment.
  • the duration corresponding to the at least part of the multiple different moments is equal to the duration corresponding to the at least another part of the multiple different moments.
  • the number of attitude data obtained based on the measurement data of the attitude sensor of the imaging device at different times is greater than 2, and the processor may specifically perform the following operations: perform multiple operations on the attitude data obtained based on the measurement data of the attitude sensor of the imaging device at different times.
  • the cutoff frequency of any sub-low-pass filtering except the cut-off frequency of the last sub-low-pass filtering is greater than the cut-off frequency of the low-pass filtering.
  • At least part of the input data of the sub-low-pass filtering is a part of the attitude data obtained based on the measurement data of the attitude sensors of the imaging devices at different times, and at least another part of the input data of the sub-low-pass filtering is the output data of other sub-low-pass filtering.
  • All of the input data of the sub-low-pass filtering in the at least another part of the sub-low-pass filtering are the output data of the at least part of the sub-low-pass filtering.
  • the processor may also perform the following operations: each time the attitude data obtained based on the measurement data of the attitude sensor of the preset number of imaging devices is acquired, perform corresponding sub-low-pass filtering; or each time the imaging device corresponding to the preset duration is acquired When the attitude data is obtained based on the measurement data of the attitude sensor, the corresponding sub-low-pass filtering is performed.
  • the imaging device is carried on the movable platform through the pan/tilt, and the pan/tilt is used to stabilize the imaging device.
  • the present embodiment also provides a movable platform carrying an imaging device for capturing images.
  • the movable platform further includes: a memory for storing executable instructions; a processor for executing the executable instructions stored in the memory to perform the following operations: acquiring motion information of the imaging device at the current capture moment of the image and based on the gesture sensor The gesture data obtained from the measurement data of Stabilize the image.
  • the movable platform may further include a pan-tilt, the movable platform carries the imaging device through the pan-tilt, and the pan-tilt is used to stabilize the imaging device.
  • the present embodiment also provides an imaging device including an image sensor for capturing an image.
  • the imaging device further includes: a memory for storing executable instructions; a processor for executing the executable instructions stored in the memory to perform the following operations: acquiring motion information of the imaging device at the current capture moment of the image and gesture sensor-based motion information.
  • the attitude data obtained from the measurement data, the image is captured by the imaging device; the target attitude data for stabilizing the image is determined based on the motion information at least in part; the attitude data enhancement is obtained according to the target attitude data and the measurement data based on the attitude sensor at the current capture moment. stabilize the image.
  • the imaging device in this embodiment may be, for example, a digital video camera, a sports camera, a single-lens reflex camera, a mirrorless camera, etc., and may be used by a user by hand, or may be used by a movable platform and/or a PTZ.

Landscapes

  • Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Studio Devices (AREA)

Abstract

一种电子增稳方法、电子增稳装置、可移动平台、成像装置以及计算机可读存储介质。电子增稳方法包括：获取图像的当前捕获时刻的成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据，图像为由成像装置捕获；至少部分根据运动信息确定用于增稳图像的目标姿态数据；根据目标姿态数据以及当前捕获时刻的姿态数据对图像进行处理。以使得处理后的图像的姿态满足用户期望。

Description

电子增稳方法及装置、可移动平台、成像装置 技术领域
本申请涉及图像处理技术领域，更具体地涉及一种电子增稳方法、电子增稳装置、可移动平台、成像装置以及计算机可读存储介质。
背景技术
通过对图像进行电子增稳处理，可以实现对图像的姿态的调整，从而消除图像拍摄过程中用户不期望的运动(例如，抖动等)对图像的姿态的影响。然而，对图像进行电子增稳处理的过程中，难以使得电子增稳处理后的图像的姿态满足用户期望。
发明内容
鉴于上述问题，提出了一种克服上述问题或者至少部分地解决上述问题的电子增稳方法、电子增稳装置、可移动平台、成像装置以及计算机可读存储介质。
根据本申请的第一个方面，提供了一种图像的电子增稳方法，包括：获取所述图像的当前捕获时刻的成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据，所述图像为由所述成像装置捕获；至少部分根据所述运动信息确定用于增稳所述图像的目标姿态数据；根据所述目标姿态数据以及所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据处理所述图像。
根据本申请的第二个方面，提供了一种电子增稳装置，所述电子增稳装置包括：存储器，用于存储可执行指令；处理器，用于执行所述存储器中存储的所述可执行指令，以执行如下操作：获取图像的当前捕获时刻的成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据，所述图像为由所述成像装置捕获；至少部分根据所述当前运动信确定用于增稳所述图像的目标姿态数据；根据所述目标姿态数据以及所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据处理所述图像。
根据本申请的第三个方面，提供了一种可移动平台，所述可移动平台承载有成像装置，所述成像装置用于捕获图像；所述可移动平台还包括：存储器，用于存储可执行指令；处理器，用于执行所述存储器中存储的所述可执行指令，以执行如下操作：获取图像的当前捕获时刻的成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据；至少部分根据所述运动信息确定用于增稳所述图像的目标姿态数据；根据所述目标姿态数据以及所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据处理所述图像。
根据本申请的第四个方面，提供了一种成像装置，所述成像装置包括图像传感器，所述图像传感器用于采集图像，其特征在于，所述成像装置还包括：存储器，用于存储可执行指令；处理器，用于执行所述存储器中存储的所述可执行指 令，以执行如下操作：获取图像的当前捕获时刻的成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据；至少部分根据所述运动信息确定用于增稳所述图像的目标姿态数据；根据所述目标姿态数据以及所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据对所述图像进行处理。
根据本申请的第五个方面，提供了一种计算机可读存储介质，其存储有可执行指令，所述可执行指令在由一个或多个处理器执行时，使所述一个或多个处理器执行本申请的第一个方面提供的所述电子增稳方法。
本申请中，图像的捕获时刻的成像装置的运动信息可以在一定程度上反映用户期望的图像的姿态，而用于对图像处理以达到电子增稳的目标姿态数据是根据图像捕获时刻的成像装置的运动信息确定的，因此，确定的目标姿态数据能够体现用户期望的图像的姿态。由此，本申请通过对图像进行处理以达到电子增稳，能够使得处理后的图像的姿态满足用户期望。
本申请的附加方面和优点将在下面的描述部分中变得明显，或通过本申请的实践了解到。本申请内容中提供的仅仅是一个实施例，而不是本申请本身，本申请内容的效果仅仅是实施例的效果，而不是本申请所有的、全部的技术效果。
附图说明
通过下文中参照附图对本申请所作的描述，本申请的其它目的和优点将显而易见，并可帮助对本申请有全面的理解。其中：
图1是根据本申请的一个实施例的可移动平台的结构示意图；
图2是根据本申请的一个实施例的电子增稳方法的目标姿态数据以及测量姿态数据的示意图；
图3是根据本申请的一个实施例的图像在电子增稳方法处理前后的第一种示意图；
图4是根据本申请的一个实施例的图像在电子增稳方法处理前后的第二种示意图；
图5是根据本申请的一个实施例的图像在电子增稳方法处理前后的第三种示意图；
图6是根据本申请的一个实施例的云台的控制原理图；
图7是根据本申请的一个实施例的图像在电子增稳方法处理前后的第四种示意图；
图8是图7所示示意图在缩减图像B的尺寸后的示意图；
图9是根据本申请的一个实施例的电子增稳方法的撞限位的原理图；
图10是根据本申请的一个实施例的电子增稳方法中的测量姿态数据对增稳效果的影响的原理图；
图11是根据本申请的一个实施例的电子增稳方法中的一种低通滤波的原理图；
图12是根据本申请的一个实施例的电子增稳方法中的另一种低通滤波的 原理图。
应该注意的是，附图并未按比例绘制，并且出于说明目的，在整个附图中类似结构或功能的元素通常用类似的附图标记来表示。还应该注意的是，附图只是为了便于描述优选实施例，而不是本申请本身。附图没有示出所描述的实施例的每个方面，并且不限制本申请的范围。
具体实施方式
下面详细描述本申请的实施方式，所述实施方式的示例在附图中示出，其中自始至终相同或类似的标号表示相同或类似的元件或具有相同或类似功能的元件。下面通过参考附图描述的实施方式是示例性的，仅用于解释本申请，而不能理解为对本申请的限制。
在本申请的描述中，需要理解的是，术语“第一”、“第二”仅用于描述目的，而不能理解为指示或暗示相对重要性或者隐含指明所指示的技术特征的数量。由此，限定有“第一”、“第二”的特征可以明示或者隐含地包括但不限于一个或者更多个所述特征。
下文的公开提供了许多不同的实施方式或例子用来实现本申请。为了简化本申请的公开，下文中对特定例子的部件和方法进行描述。当然，它们仅仅为示例，并且目的不在于限制本申请。此外，本申请可以在不同例子中重复参考数字和/或参考字母，这种重复是为了简化和清楚的目的，其本身不指示所讨论各种实施方式和/或设置之间的关系。
本实施例，首先提供了一种图像的电子增稳方法，图1是根据本申请的一个实施例的可移动平台10的结构示意图，该可移动平台10以无人机为例。如图1所示，图像为由成像装置100捕获，成像装置100承载于可移动平台10。
在一些实施例中，成像装置100可以为照相机、摄像机等，具体地，可以为数码摄像机、运动相机、单反相机、微单相机等，也可以为其他各种航拍设备。在另一些实施例中，成像装置可以为具有拍摄功能的智能终端，例如具有拍摄功能的手机、平板等。
其中，图1中可移动平台10以无人机为例，可以理解地，无人机通常也被称为UAV(Unmanned Aerial Vehicle，无人飞行器)，其中，无人机可以包括固定翼无人机、旋翼无人机、伞翼无人机等各种类型。可以理解地，成像装置100与无人机的机身的连接处并不限于图1所示的位置，也就是说，成像装置100不仅可以与无人机的底部连接，也可以与无人机的顶部、侧部等位置连接，本实施例对此并不加以限制。
在其他实施例中，可移动平台还可以为手持云台、无人车、机器人等。
当可移动平台为手持云台时，手持云台可以包括手柄和云台，云台用于承载成像装置。可以理解地，手柄可以用于支撑云台，且手柄的形状并不限于柱状，也就是说，手柄不仅可以为圆柱状、棱柱状等，还可以为圆台状、棱锥状、球状等，甚至是上述各种形状的组合或异形形状等，本实施例对手柄的具体形状并不 加以限定。这种可移动平台可以直接被用户手持操作，应用场景广泛，便于用户操作，且节省成本。
可以理解地，手柄上可以设置有操作装置，操作装置可以为操作按键、操作杆或操控界面等，以便于控制成像装置或可移动平台的其他部件，例如操控成像装置的开启、关闭以及拍摄等。
当可移动平台为无人车时，无人车可以包括底盘。可以理解地，底盘可以用于支撑成像装置，且底盘的移动方式并不限于利用轮子进行移动，也可以通过履带等其他机构移动。其中，当无人车直接利用轮子进行移动时，无人车的轮子的数量可以为一个或多个，本实施例对此并不加以限定。
当可移动平台为机器人时，机器人可以包括机身。可以理解地，成像装置与机器人的机身的连接处并不限于机器人的机身的机器头部连接，也可以与器人的机身的机器手臂、机器背部等其他部位连接，本实施例对此并不加以限定。
本实施例提供的图像的电子增稳方法包括：获取图像的当前捕获时刻的成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据，图像为由成像装置捕获；至少部分根据当前运动信度确定用于增稳图像的目标姿态数据；根据目标姿态数据以及当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据处理图像。
其中，当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据为通过姿态传感器的测量数据直接获得或融合得到，姿态传感器包括但不限于惯性测量单元。成像装置的运动信息可以包括移动角速度、移动角加速度中的一种，也可以通过各种传感器获得，如陀螺仪或加速度计，还可以通过该捕获时刻的前后帧的图像的角速度信息估计获得，本实施例对此并不加以限制。
其中，运动信息还可以包括平移速度、平移加速度中的一种，该平移速度或平移加速度可以通过线运动传感器得到，如全球定位系统GPS。在运动信息还包括平移速度、平移加速度中的一种时，还可以获取图像的当前捕获时刻的成像装置的位置数据，以进一步结合当前捕获时刻的位置数据对图像进行处理，达到电子增稳的目的。可以理解，移动角速度以及移动角速度主要考虑到成像装置的旋转运动对成像质量的影响，平移速度以及平移加速度主要考虑到成像装置的平移运动对成像质量的影响。本申请实施例下述内容以像装置的旋转运动对成像质量的影响进行示例性说明。
图2是根据本申请的一个实施例的电子增稳方法的目标姿态数据b(图中虚线)以及基于姿态传感器的测量数据获得的姿态数据a(图中实线)的示意图，如图2所示，其中横坐标t表示时间，纵坐标v表示姿态，其中，图2以目标姿态数据在图2所示的坐标系中为正弦形为例，但可以理解地，在其他实施例中，目标姿态数据在图2所示的坐标系中可以为其他各种形状，其可以是与用户对可移动平台和/或云台的控制相关，其可以是对基于姿态传感器的测量数据获得的姿态数据进行平滑处理得到。其中，图2以目标姿态数据b以及基于姿态传感器的测量数据获得的姿态数据a仅在t1时刻附近存在差异，在其他时刻均是一致 的情况作为示例。
在成像装置承载于可移动平台并捕获图像时，成像装置由于可移动平台或云台的原因，可能会受到诸如风力、气压变化等各种因素的影响而产生用户不期望的运动，例如抖动，例如，图2中t1表示的时刻，此时，会导致成像装置捕获图像时的姿态并不是用户期望的。而在图2中t2表示的时刻时，目标姿态数据以及基于姿态传感器的测量数据获得的姿态数据是一致的，此时，成像装置捕获图像时的姿态是用户期望的。
其中，可以理解的是，成像装置也可以不承载于可移动平台而独立使用，在这种情况下，当成像装置由用户手持操作时，也可能由于用户的手部抖动而产生用户不期望的运动。
具体的，根据目标姿态数据以及当前捕获时刻的姿态数据处理图像，可以是根据目标姿态数据以及当前捕获时刻的姿态数据计算该时刻的姿态补偿量。然后，根据姿态补偿量对电子增稳处理前的图像进行校正，以得到处理后的图像。
其中，若获取图像的当前捕获时刻的成像装置的位置数据，还可以根据目标位置数据以及当前捕获时刻的位置数据计算该时刻的位置补偿量，以根据位置补偿量对电子增稳处理前的图像进行校正，以得到处理后的图像。
其中，为了避免处理后的图像包括偏离图像的区域，上述校正可以通过对电子增稳处理前的图像进行裁剪实现。上述校正可以包括旋转校正和/或平移校正，旋转校正即在图像上旋转裁剪区域，平移校正即在图像的上下、左右方向上移动裁剪区域。
可以理解的是，在考虑成像装置的旋转运动对成像质量的影响时，上述校正可以包括旋转校正和/或平移校正；在考虑成像装置的平移运动对成像质量的影响时，上述校正可以包括平移校正。
其中，若目标姿态数据以及基于姿态传感器的测量数据获得的姿态数据是一致的，则处理后的图像位于处理前的图像的预设位置处。图3是根据本申请的一个实施例的图像在电子增稳方法处理前后的第一种示意图。其中，图像A表示电子增稳方法处理前的图像，图像B表示电子增稳方法处理前的图像。如图3所示，在一些实施例中，预设位置可以位于图像A的中部，由此，保证在目标姿态数据以及基于姿态传感器的测量数据获得的姿态数据不一致时，图像A可以在各个方向上均具有适合的裁剪空间。在其他实施例中，预设位置也可以不位于图像A的中部。
可以理解地，根据姿态补偿量对电子增稳处理前的图像进行裁剪可以是：使得裁剪后的图像在裁剪前的图像中的位置偏离预设位置的方向，与对应时刻的基于姿态传感器的测量数据获得的姿态偏离目标姿态的方向相反。
例如，图4是根据本申请的一个实施例的图像在电子增稳方法处理前后的第二种示意图。图4对应的图像A的当前捕获时刻的基于姿态传感器的测量数据获得的姿态偏离目标姿态的方向为向上，则使得裁剪后的图像在裁剪前的图像中的位置偏离预设位置的方向为向下。
可以理解地，若图像A的当前捕获时刻的基于姿态传感器的测量数据获得的姿态偏离目标姿态的方向为向下，则使得裁剪后的图像在裁剪前的图像中的位置偏离预设位置的方向为向上。
又例如，图5是根据本申请的一个实施例的图像在电子增稳方法处理前后的第三种示意图。图5对应的图像A的当前捕获时刻的测量姿态偏离目标姿态的方向为逆时针旋转，则使得裁剪后的图像在裁剪前的图像中的位置偏离预设位置的方向为顺时针旋转。
可以理解地，若图像A的当前捕获时刻的测量姿态偏离目标姿态的方向为顺时针旋转，则使得裁剪后的图像在裁剪前的图像中的位置偏离预设位置的方向为逆时针旋转。
也就是说，当目标姿态数据与用户期望的成像装置的移动对应，就可以根据目标姿态数据对图像进行处理以达到电子增稳，并得到具有用户期望的姿态的图像。
本实施例中，图像的捕获时刻的成像装置的运动信息可以在一定程度上反映用户期望的图像的姿态，而对图像进行处理以达到电子增稳的目标姿态数据是根据图像捕获时刻的成像装置的运动信息确定的，因此，确定的目标姿态数据能够体现用户期望的图像的姿态。由此，本申请通过对图像进行处理以达到电子增稳处理，能够使得处理后的图像的姿态满足用户期望。
在一些实施例中，根据移动角速度确定用于增稳图像的目标姿态数据可以包括：根据移动角速度与角速度阈值的比较结果，确定用于增稳图像的目标姿态数据。
其中，由于移动角速度的具体数值可以多种多样，而移动角速度与角速度阈值的比较结果的情况相对有限，因此，根据移动角速度与角速度阈值的比较结果，确定用于增稳图像的目标姿态数据可以避免移动角速度的具体数值多种多样带来的各种复杂情况，并使得确定目标姿态数据的过程简单。下面针对成像装置的移动角速度对图像增稳的影响、以及处理策略进行具体说明：
电子增稳方法还可以包括：获取姿态控制指令，姿态控制指令用于控制成像装置的目标姿态。则根据移动角速度与角速度阈值的比较结果，确定用于增稳图像的目标姿态数据可以包括：当移动角速度小于或等于角速度阈值时(此时，可移动平台可以处于悬停状态或低速运动状态)，确定用于增稳图像的目标姿态数据为与姿态控制指令对应的成像装置的目标姿态。
可以理解地，姿态控制指令对应的成像装置的目标姿态可以表示用户期望的成像装置的目标姿态，因此，通过确定用于增稳图像的目标姿态数据为与姿态控制指令对应的成像装置的目标姿态，可以得到具有用户期望的姿态的图像，即通过电子增稳方法处理后可以消除所有图像拍摄过程中用户不期望的运动对图像的姿态的影响，提升用户体验。这种确定用于增稳图像的目标姿态数据的方式，相比于通过滤波器对测量姿态进行处理以得到用于增稳图像的目标姿态数据的方式，可以相当于实现对不同频段的噪声进行消除，即使在对画面要求比较高的 情况下，例如高倍率变焦的情况下，电子增稳处理后的图像也不会有人眼可察觉到的抖动。
可以理解地，姿态控制指令可以由用户通过各种方式发出，例如，通过可移动平台的控制终端，如遥控器、各种智能设备等发出。当姿态控制指令由用户通过可移动平台的遥控器发出时，可以通过遥控器上的按钮、摇杆、触摸屏等发出。当姿态控制指令由用户通过智能设备发出时，可以通过智能设备上的APP发出，其中，智能设备可以为手机、平板、电脑等。
在一些实施例中，由于成像装置承载于可移动平台上，姿态控制指令用于指示通过控制可移动平台的姿态来实现成像装置的目标姿态的控制。
在另一些实施例中，姿态控制指令用于指示通过控制云台(如图1所示，可移动平台可以包括云台200)的姿态来实现成像装置的目标姿态的控制，成像装置通过云台承载于可移动平台。
其中，云台可以包括一个云台部件、两个云台部件、三个云台部件或更多个云台部件，相应地，云台可以允许成像装置绕一个、两个、三个或更多个轴旋转，用于旋转的轴可以彼此正交，也可以不是正交。在一些实施例中，云台部件通过电机可以控制成像装置的姿态，包括控制成像装置的俯仰角、横滚角以及偏航角中的一个或多个，相应地，成像装置可以绕俯仰轴、横滚轴以及偏航轴中的一个或多个旋转。
在一些实施例中，云台部件可以为3个，如第一云台部件、第二云台部件以及第三云台部件，可以理解地，每个云台部件可以包括连接臂。其中，示例性的，第一云台部件与可移动平台的本体连接，并且第一云台部件可以相对本体转动，以使得成像装置的偏航角发生变化，即第一连接臂相对本体转动时，可以使得成像装置绕偏航轴旋转。第二云台部件与第一云台部件连接，并且第二云台部件可以相对本体转动，以使得成像装置的横滚角发生变化，即第二云台部件相对本体转动时，可以使得成像装置绕横滚轴旋转。第三云台部件与第二云台部件连接，并且第三云台部件可以相对本体转动，以使得成像装置的俯仰角发生变化，即第三云台部件相对本体转动时，可以使得成像装置绕俯仰轴旋转。
在另一些实施例中，云台可以仅包括一个云台部件，这个云台部件可以相对本体转动，以使得成像装置的偏航角发生变化，即这个云台部件相对本体转动时，可以使得成像装置绕偏航轴旋转。
可以理解地，当可移动平台为无人机时，与本体连接的云台可以有一个云台部件、两个云台部件、三个云台部件或更多个云台部件，并且可以使得成像装置绕俯仰轴、横滚轴以及偏航轴中的一个、两个、三个旋转，从而使得成像装置也可以绕更多的轴旋转等。可移动平台为手持云台、机器人或无人车时，云台也可以有一个云台部件、两个云台部件、三个或三个以上的云台部件，并且可以使得成像装置绕俯仰轴、横滚轴以及偏航轴中的一个或、两个或三个旋转，从而使得成像装置也可以绕三个以上的轴旋转等。也就是说，不论可移动平台是什么类型，云台都既可以为单轴云台、双轴云台、三轴云台或其他轴数的云台。
图6是根据本申请的一个实施例的云台的控制原理图。如图6所示，云台在调整成像装置的姿态时，是通过惯性测量元件检测成像装置的当前姿态，并将成像装置的当前姿态并和目标姿态做比较，求出控制偏差，控制系统根据控制偏差控制电机，以调整成像装置的姿态，最终减小控制偏差，保证成像装置的实际姿态和目标姿态偏差尽量小。
本实施例可以将用于实现机械增稳的云台与用于对图像进行处理的电子增稳方法结合，从而避免云台由于云台的控制精度、云台的惯性测量元件的测量精度造成的其增稳效果不稳定等问题的发生，通过使得机械增稳的云台与能够实现无误差控制的电子增稳方法相结合，使得即使在对画面要求比较高的情况下，例如高倍率变焦的情况下，这种机械增稳与电子增稳方法相结合的方式得到的图像也不会有人眼可察觉到的抖动。
可以理解地，根据姿态补偿量对电子增稳处理前的图像进行裁剪可以是：使得裁剪后的图像在裁剪前的图像中的位置偏离预设位置的程度，与，对应时刻的测量姿态偏离目标姿态的程度成正比。即测量姿态偏离目标姿态的程度越大，则裁剪后的图像在裁剪前的图像中的位置偏离预设位置的程度越大。由此，可以保证处理后的图像是用户期望的。
图7是根据本申请的一个实施例的图像在电子增稳方法处理前后的第四种示意图。图7对应的图像A的捕获时刻的测量姿态偏离目标姿态的方向为逆时针旋转，则使得裁剪后的图像在裁剪前的图像中的位置偏离预设位置的方向为顺时针旋转。当测量姿态偏离目标姿态的程度较大时，可能会出现图7中所示的区域C，而图像A中是不具有区域C的，因此，此时会使得处理后的图像B出现用户不期望的区域C。而在移动角速度大于角速度阈值(此时，可移动平台可以处于大机动运动)时，会由于可移动平台或云台响应用户输出的姿态控制指令存在的滞后性，而导致测量姿态偏离目标姿态的程度较大，容易使得处理后的图像B出现上述区域C。
因此，本实施例中，根据移动角速度与角速度阈值的比较结果，确定用于增稳图像的目标姿态数据可以包括：当移动角速度大于角速度阈值(此时，可移动平台可以处于大机动运动)时，根据当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据以及预设偏置阈值确定用于增稳图像的目标姿态数据。从而避免处理后的图像B出现用户不期望的区域C，以提升用户体验。其中，预设偏置阈值用于保证成像装置拍摄的图像中内容变化的连续性，以降低外部扰动对成像装置的成像质量的影响。
在一些实施例中，角速度阈值可以与图像在处理前后的尺寸差异相关。由于角速度阈值与避免处理后的图像B出现用户不期望的区域C有关，而在测量姿态偏离目标姿态的程度相同的情况下，处理后的图像B是否出现用户不期望的区域C与图像在处理前后的尺寸差异相关，例如，请同时参阅图7以及图8，图8是图7所示示意图在缩减图像B的尺寸后的示意图。其中，在基于姿态传感器的测量数据获得的姿态偏离目标姿态的程度相同的情况下，处理后的图像B可以出现 用户不期望的区域C(如图7)，也可以不出现用户不期望的区域C(如图8)，从图7以及图8可以看出在基于姿态传感器的测量数据获得的姿态偏离目标姿态的程度相同的情况下，处理后的图像B是否出现用户不期望的区域C与图像在处理前后的尺寸差异相关。因此，通过使得角速度阈值与图像在处理前后的尺寸差异相关，可以使得处理后的图像B不容易出现上述区域C。
具体地，角速度阈值与图像在处理前后的尺寸差异可以正相关。也就是说，图像在处理前后的尺寸差异越大，则角速度阈值可以越大，图像在处理前后的尺寸差异越小，则角速度阈值可以越小。如图7、图8所示，当图像在处理前后的尺寸差异越大时，当基于姿态传感器的测量数据获得的姿态姿态偏离目标姿态的程度越大才会出现上述区域C，因此，可以使得角速度阈值越大(此时，可移动平台或云台响应用户输出的姿态控制指令的滞后程度越大，测量姿态偏离目标姿态的程度越大)。由此，可以在保证处理后的图像B不出现上述区域C的同时，最大可能使得用于增稳图像的目标姿态数据为与姿态控制指令对应的成像装置的目标姿态。当用于增稳图像的目标姿态数据为与姿态控制指令对应的成像装置的目标姿态时，可以对不同频段的噪声进行消除，即使在对画面要求比较高的情况下，例如高倍率变焦的情况下，电子增稳处理后的图像也不会有人眼可察觉到的抖动。
在一些情况下，本实施例的电子增稳方法还可以包括：获取多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据；则根据移动角速度确定用于增稳图像的目标姿态数据可以包括：根据移动角速度，对多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据进行筛选；根据筛选后的姿态数据，确定用于增稳图像的目标姿态数据。
由此可知，在确定用于增稳图像的目标姿态数据时，用到了多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据，而用户不期望的成像装置的变化通常只在一小部分的时间段产生，也就是说，多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据通常大多数对应的都是用户期望的成像装置的姿态。因此，通过多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据来确定图像捕获时刻的用于增稳图像的目标姿态数据，可以在一定程度上使得用于增稳图像的目标姿态数据符合用户期望。
根据移动角速度，对多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据进行筛选可以包括：根据移动角速度确定低通滤波的截止频率；根据低通滤波的截止频率对多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据进行低通滤波，以得到筛选后的姿态数据。
低通滤波是一种过滤方式，规则为低频信号能正常通过，而超过设定临界值(截止频率)的高频信号则被阻隔、减弱。但是阻隔、减弱的幅度可以依据不同的频率以及不同的滤波程序(目的)而改变。可以理解，低通滤波有的时候也被称作高频去除过滤或者最高去除过滤。
采用滤波的方式对多个不同时刻的成像装置的基于姿态传感器的测量数据 获得的姿态数据进行处理，可以有效消除图像拍摄过程中用户不期望的运动(例如，抖动等)对图像的姿态的影响。
其中，根据移动角速度确定低通滤波的截止频率可以包括：根据移动角速度与移动角速度阈值的比较结果，确定低通滤波的截止频率。
在一些实施例中，可移动平台10的移动角速度可以通过各种传感器获得，还可以通过该捕获时刻的前后帧的图像的角角速度信息估计获得，本实施例对此并不加以限制。在另一些实施例中，可移动平台10的移动角速度可以通过用户输入的用于控制成像装置的移动角速度的指令获得，例如，当用户通过摇杆输入的用于控制成像装置的移动角速度的指令，可移动平台10的移动角速度可以通过摇杆的杆量获得。
根据移动角速度与移动角速度阈值的比较结果，确定低通滤波的截止频率可以包括：当移动角速度小于或等于移动角速度阈值时，确定低通滤波的截止频率为预设截止频率。由此，可以有效消除图像拍摄过程中用户不期望的运动(例如，抖动等)对图像的姿态的影响，保证图像的平滑效果。
根据移动角速度与移动角速度阈值的比较结果，确定低通滤波的截止频率可以包括：当移动角速度大于移动角速度阈值时，调整低通滤波的截止频率，以大于预设截止频率。
如图7所示，当测量姿态偏离目标姿态的程度较大时，可能会出现图7中所示的区域C，而图像A中是不具有区域C的，因此，此时会使得处理后的图像B出现用户不期望的区域C。而在移动角速度大于移动角速度阈值时，会由于可移动平台或云台响应用户输出的姿态控制指令存在的滞后性，而导致基于姿态传感器的测量数据获得的姿态偏离目标姿态的程度较大，容易使得处理后的图像B出现上述区域C。为了避免处理后的图像B出现上述区域C可以适当地在通过上述方法确定目标姿态后，对目标姿态进行校正，以使得根据校正后的目标姿态对图像进行处理时不会出现上述区域C。由于对图像进行顺时针旋转与校正前的目标姿态对应的角度后出现上述区域C，因此，可以在对图像进行顺时针旋转时，使其旋转的角度与校正后的目标姿态对应。可以理解地，校正前的目标姿态对应的角度大于校正后的目标姿态对应的角度，由此，可以避免图像B出像上述区域C(如图9所示)。此时，目标姿态会在一定程度上偏离用户的期望，使得图像的画面不能顺滑过渡，甚至出现抽动，这种现象通常也可以被称为撞限位，图9是根据本申请的一个实施例的电子增稳方法的撞限位的原理图。
可以理解地，当低通滤波的截止频率越低时，则测量姿态与目标姿态不一致的情况越严重，尤其是当成像装置的移动角速度越快时，基于姿态传感器的测量数据获得的姿态偏离目标姿态的程度较大的情况越严重，则越容易发生上述撞限位的现象或越容易使得处理后的图像B出现上述区域C，因此，当移动角速度大于移动角速度阈值时，调整低通滤波的截止频率，以大于预设截止频率，从而避免发生上述撞限位的现象或避免处理后的图像B出现上述区域C，以提升用户的使用体验。
当移动角速度大于移动角速度阈值的幅度越大时，低通滤波的截止频率与预设截止频率的差异幅度可以越大。由于，当成像装置的移动角速度越快时，基于姿态传感器的测量数据获得的姿态偏离目标姿态的程度较大的情况越严重，则越容易发生上述撞限位的现象或越容易使得处理后的图像B出现上述区域C，因此，当移动角速度大于移动角速度阈值的幅度越大时，低通滤波的截止频率与预设截止频率的差异幅度越大，从而避免发生上述撞限位的现象或避免处理后的图像B出现上述区域C，以提升用户的使用体验。
其中，可以理解，上述的移动角速度阈值与角速度阈值可以不同，也可以相同。在不同的情况下，上述依据移动角速度确定用于增稳图像的目标姿态数据的相应实施例之间可以耦合，在相同的情况下，上述依据移动角速度确定用于增稳图像的目标姿态数据的相应实施例可以根据需要选择使用。
进一步的，在上述用于增稳图像的目标姿态数据的相应实施例选择使用时，若成像装置通过云台搭载于可移动平台上，那么可以基于移动角速度与角速度阈值的比较，确定用于增稳图像的目标姿态数据，相对而言，该方法可以利用姿态控制指令实现对不同频段的噪声进行消除；若成像装置独立应用或直接承载于可移动平台，那么可以基于移动角速度与移动角速度阈值的比较，确定用于增稳图像的目标姿态数据，以在没有云台承载的情况下，仍能够较好地实现图像增稳。
当然，尽管在成像装置承载于可移动平台上时，成像装置的目标姿态也可以通过可移动平台的姿态控制来实现，但在该应用场景下，可移动平台与成像装置固连，可移动平台更多实现的是轨迹的变化，而非主要用于对成像装置的姿态调整，而云台更多实现的是对成像装置的姿态调整，因此，在利用姿态控制指令确定用于增稳图像的目标姿态数据时，可以主要考虑成像装置经云台搭载于可移动平台的情况。
需要说明的是，在运动信息包括移动角加速度时，可以利用移动角加速度积分得到的移动角速度进行与角速度阈值、移动角速度阈值的比较。或者，也可以是，直接对移动角加速度与设定阈值进行比较。具体判断方法、判断结果与前述内容相似，此处不再赘述。
在一些实施例中，可移动平台可以设置有一个截止频率可变的滤波器，以通过调整该滤波器的截止频率实现对低通滤波的截止频率的调整。在另一些实施例中，可移动平台可以设置有多个不同截止频率的滤波器，在低通滤波时选择移动角速度对应的滤波器进行低通滤波，或者对不同滤波器的低通滤波结果进行加权平均，以得到与根据移动角速度确定的截止频率对应的目标姿态数据。
多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据可以包括：当前捕获时刻的姿态数据。图像的当前捕获时刻的成像装置的测量姿态数据在能一定程度上体现该时刻用户期望的成像装置的姿态，使得多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据包括该数据可以使得得到的目标姿态数据符合用户期望。
多个不同时刻中的至少部分位于当前捕获时刻之前，多个不同时刻中的至少 另一部分位于当前捕获时刻之后。多个不同时刻中的该至少部分对应的时长，与多个不同时刻中的该至少另一部分对应的时长可以相等。
由此，保证对图像的增稳效果，以成像装置单调运动为例，其中单调运动是指成像装置朝向一个方向运动，包括匀速、加速、减速等。以匀速运动为例，图10是根据本申请的一个实施例的电子增稳方法中的基于姿态传感器的测量数据获得的姿态数据对增稳效果的影响的原理图，参见图10，成像装置的基于姿态传感器的测量数据获得的姿态如图10中(a)图所示，其中图10中(a)图中实际曝光点ET，捕获时刻T0，基于姿态传感器的测量数据获得的姿态数据包括捕获时刻T0之前第一时间段T1内的第一姿态数据，以及捕获时刻T0之后第二时间段T2内的第二姿态数据。
参见图10中(b)图，其中采集第一姿态数据等工作所用时间为Delta-t1，第一姿态数据对应实际姿态IT，基于该第一姿态数据可以得到目标姿态IE。在采用频域低通滤波器为中值滤波器的情况下，第一姿态数据的中值点EE位于T1/2处。由于后续对实际曝光点ET处图像采用中值点EE处的数据增稳，因此在中值点EE和实际曝光点ET之间存在延时Delta-t2，其中Delta-t2等于T1/2。该延时是由于低通滤波器所得到的处理结果即中值点EE与实际曝光点ET的偏差，或者可理解为滤波误差，若成像装置在延时Delta-t2期间发生抖动，采用中值点EE处的数据对待增稳图像进行增稳会发生偏差。
参见图10中(c)图，其中第一姿态数据和第二姿态数据采集等工作所用时间为Delta-t1，第一姿态数据和第二姿态数据对应实际姿态IT，以T1和T2相等为例，基于第一姿态数据和第二姿态数据可以得到目标姿态IE。在采用频域低通滤波器为中值滤波器的情况下，第一姿态数据和第二姿态数据的中值点EE位于T0处，即与实际曝光点ET重合，从而可以避免出现(b)图中所存在的T1/2的延时Delta-t2即延时Delta-t2等于0。这样采用中值点EE处的数据对图像进行增稳的结果是比较准确的。
由此可见，多个不同时刻中的至少部分位于当前捕获时刻之前，多个不同时刻中的至少另一部分位于当前捕获时刻之后，多个不同时刻中的该至少部分对应的时长，与多个不同时刻中的该至少另一部分对应的时长可以相等的设置可以保证对图像的增稳效果。
并且，可以弥补第一姿态数据和第二姿态数据采集等工作造成的低通滤波的延迟对滤波结果的影响，从而可以使得低通滤波具有较低的截止频率以对更多地数据进行处理，可以理解地，当截止频率越低时，图像的画面就越顺滑。
不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据的数量可以大于2，以提高增稳的效果。可以理解地，在确定用于增稳图像的目标姿态数据时用到了多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据，而用户不期望的成像装置的变化通常只在一小部分的时间段产生，也就是说，多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据通常大多数对应的都是用户期望的成像装置的姿态，因此，通过多个不同时刻的成 像装置的基于姿态传感器的测量数据获得的姿态数据来确定图像当前捕获时刻的用于增稳图像的目标姿态数据可以在一定程度上使得用于增稳图像的目标姿态数据符合用户期望，而不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据越多，就越能消除捕获时刻不期望的运动对结果的影响，也就是说，当不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据越多时，处理后的图像的画面越顺滑，电子增稳的效果越好。
根据低通滤波的截止频率对多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据进行低通滤波可以包括：对多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据进行多次子低通滤波。
由此，使得低通滤波的计算过程可以分别进行，避免由于计算集中造成计算资源浪费的现象，可以缓解计算压力。并且，在滤波过程中，仍然采用了多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据的全部，因此，可以达到与对多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据进行一次滤波相同或类似的效果。
其中，最后一次子低通滤波(可以为输出数据与该多个不同时刻的测量姿态数据均相关的子低通滤波，即最后一次子低通滤波输出数据是该多个不同时刻的测量姿态数据中所有的都经过低通滤波的结果)的截止频率为根据低通滤波的截止频率确定。由此，保证得到的目标数据与根据移动角速度确定的低通滤波的截止频率对应。
在一些实施例中，除最后一次子低通滤波的截止频率以外的任何一次子低通滤波的截止频率，大于低通滤波的截止频率。从而使得最终获得的图像能够保留用户期望的运动对图像的姿态的影响。
在任一次子低通滤波输出对应的输出数据后，该任一次子低通滤波的输入数据被删除。由此，可以避免数据占用存储空间大的问题，缓解存储压力，在任一次子低通滤波输出对应的输出数据后，该任一次子低通滤波的输入数据被删除，即在相应的计算过程中是存储有该任一次子低通滤波的输入数据的，从而在计算出现事故需要重新获得对应的输入数据时，也能立刻获得对应的输入数据。
在其他实施例中，可以在该任一次子低通滤波输出对应的输出数据的过程中，或者在该任一次子低通滤波正在进行计算时，该任一次子低通滤波的输入数据被删除。
图11是根据本申请的一个实施例的电子增稳方法中的一种低通滤波的原理图，其中，d1、d2、d3、d4、d5、d6、d7以及d8分别对应一个时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据，f1、f2、f3、f4、f5、f6以及f7分别表示不同子低通滤波。如图11所示，至少部分子低通滤波(f1、f2、f3以及f4表示的子低通滤波)的输入数据为多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据(d1、d2、d3、d4、d5、d6、d7以及d8对应的成像装置的基于姿态传感器的测量数据获得的姿态数据)的部分，至少另一部分子低通滤波(f5、f6以及f7表示的子低通滤波)的输入数据为其他子低通滤 波的输出数据。
可以将输入数据为多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据的子低通滤波称为一级滤波(f1、f2、f3以及f4表示的子低通滤波)，将输入数据为一级滤波的输出数据的子低通滤波称为二级滤波(f5、以及f6表示的子低通滤波)，将输入数据为二级滤波的输出数据的子低通滤波称为三级滤波(f7表示的子低通滤波)，并以此类推。
可以理解地，子低通滤波的次数越多，则缓解计算压力以及存储压力的效果越突出。
在一些实施例中，该至少另一部分子低通滤波中的所有的子低通滤波的输入数据为该至少部分子低通滤波的输出数据。例如，如图12所示，图12是根据本申请的一个实施例的电子增稳方法中的另一种低通滤波的原理图。即低通滤波仅包括上述一级滤波(f1、f2、f3以及f4表示的子低通滤波)和二级滤波(f5表示的子低通滤波)，由此，简化低通滤波过程。
对该多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据进行多次子低通滤波可以包括：每获取到预设个数的成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的子低通滤波。例如，每获取到2个、3个、4个、5个、6个、7个、8个、9个、10个、11个、12个、13个、14个、15个、16个、17个、18个、19个、20个、21个、22个、23个、24个、25个、26个或更多个成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的子低通滤波，由此，便于二级滤波的计算。
对该多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据进行多次子低通滤波还可以包括：每获取到预设时长对应的成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的子低通滤波。预设时长可以为0.1s、0.2s、0.3s、0.4s、0.5s、0.6s、0.7s、0.8s、0.9s、1.0s、1.1s、1.2s、1.3s、1.4s、1.5s、1.6s、1.7s、1.8s、1.9s、2.0s、2.1s、2.2s、2.3s、2.4s、2.5s等。
以预设时长为0.1s，该多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据对应的总时长为2s，且低通滤波仅包括一级滤波和二级滤波为例进行说明。
低通滤波的具体过程可以包括：利用第一个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第一次一级滤波，第一次一级滤波输出结果后，删除第一次一级滤波的输入数据。利用第二个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第二次一级滤波，第二次一级滤波输出结果后，删除第二次一级滤波的输入数据。利用第三个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第三次一级滤波，第三次一级滤波输出结果后，删除第三次一级滤波的输入数据。利用第四个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第四次一级滤波，第四次一级滤波输出结果后，删除第四次一级滤波的输入数据。利用第五个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行 第五次一级滤波，第五次一级滤波输出结果后，删除第五次一级滤波的输入数据。利用第六个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第六次一级滤波，第六次一级滤波输出结果后，删除第六次一级滤波的输入数据。利用第七个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第七次一级滤波，第七次一级滤波输出结果后，删除第七次一级滤波的输入数据。利用第八个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第八次一级滤波，第八次一级滤波输出结果后，删除第八次一级滤波的输入数据。利用第九个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第九次一级滤波，第九次一级滤波输出结果后，删除第九次一级滤波的输入数据。利用第十个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第十次一级滤波，第十次一级滤波输出结果后，删除第十次一级滤波的输入数据。利用第十一个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第十一次一级滤波，第十一次一级滤波输出结果后，删除第十一次一级滤波的输入数据。利用第十二个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第十二次一级滤波，第十二次一级滤波输出结果后，删除第十二次一级滤波的输入数据。利用第十三个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第十三次一级滤波，第十三次一级滤波输出结果后，删除第十三次一级滤波的输入数据。利用第十四个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第十四次一级滤波，第十四次一级滤波输出结果后，删除第十四次一级滤波的输入数据。利用第十五个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第十五次一级滤波，第十五次一级滤波输出结果后，删除第十五次一级滤波的输入数据。利用第十六个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第十六次一级滤波，第十六次一级滤波输出结果后，删除第十六次一级滤波的输入数据。利用第十七个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第十七次一级滤波，第十七次一级滤波输出结果后，删除第十七次一级滤波的输入数据。利用第十八个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第十八次一级滤波，第十八次一级滤波输出结果后，删除第十八次一级滤波的输入数据。利用第十九个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第十九次一级滤波，第十九次一级滤波输出结果后，删除第十九次一级滤波的输入数据。利用第二十个0.1s内的基于姿态传感器的测量数据获得的姿态数据进行第二十次一级滤波，第二十次一级滤波输出结果后，删除第二十次一级滤波的输入数据。然后，利用所有一级滤波的输出结果进行二级滤波。
在一些实施例中，该多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据对应的总时长可以长于图像对应的时长(图像可以为视频)，例如，是图像对应的时长的2倍、3倍、4倍、5倍等，从而保证对图像的增稳效果。
成像装置可以通过云台承载于可移动平台，云台用于增稳成像装置。本实施例可以将用于实现机械增稳的云台与用于对图像进行处理的电子增稳方法结合， 从而避免云台由于云台的控制精度、云台的惯性测量元件的测量精度造成的其增稳效果不稳定等问题的发生，通过使得机械增稳的云台与能够实现无误差控制的电子增稳方法相结合，使得即使在对画面要求比较高的情况下，例如高倍率变焦的情况下，这种机械增稳与电子增稳方法相结合的方式得到的图像也不会有人眼可察觉到的抖动。
本实施例还提供了一种计算机可读存储介质，计算机可读存储介质存储有可执行指令，可执行指令在由一个或多个处理器执行时，可以使一个或多个处理器执行上述任一电子增稳方法。
其中计算机可读存储介质也可以被称为存储器，可执行指令又可以被称为程序。处理器可以根据存储在只读存储器(ROM)中的程序或者加载到随机访问存储器(RAM)中的程序而执行各种适当的动作和处理。处理器例如可以包括通用微处理器(例如CPU)、指令集处理器和/或相关芯片组和/或专用微处理器(例如，专用集成电路(ASIC))，等等。处理器还可以包括用于缓存用途的板载存储器。处理器可以包括用于执行根据本实施例的方法流程的不同动作的单一处理单元或者是多个处理单元。
处理器、ROM以及RAM通过总线彼此相连。处理器通过执行ROM和/或RAM中的程序来执行根据本实施例的方法流程的各种操作。需要注意，程序也可以存储在除ROM和RAM以外的一个或多个存储器中。处理器也可以通过执行存储在一个或多个存储器中的程序来执行根据本实施例的方法流程的各种操作。
根据本实施例，应用计算机可读存储介质的装置还可以包括输入/输出(I/O)接口，输入/输出(I/O)接口也连接至总线。应用计算机可读存储介质的装置还可以包括连接至I/O接口的以下部件中的一项或多项：包括键盘、鼠标等的输入部分；包括诸如阴极射线管(CRT)、液晶显示器(LCD)等以及扬声器等的输出部分；包括硬盘等的存储部分；以及包括诸如LAN卡、调制解调器等的网络接口卡的通信部分。通信部分经由诸如因特网的网络执行通信处理。驱动器也根据需要连接至I/O接口。可拆卸介质，诸如磁盘、光盘、磁光盘、半导体存储器等等，根据需要安装在驱动器上，以便于从其上读出的计算机程序根据需要被安装入存储部分。
根据本实施例的方法流程可以被实现为计算机软件程序。例如，本实施例包括一种计算机程序产品，其包括承载在计算机可读存储介质上的计算机程序，该计算机程序包含用于执行流程图所示的方法的程序代码。在这样的实施例中，该计算机程序可以通过通信部分从网络上被下载和安装，和/或从可拆卸介质被安装。在该计算机程序被处理器执行时，执行本实施例的系统中限定的上述功能。
可以理解地，计算机可读存储介质可以包括但不限于非易失性或易失性存储介质，例如随机存取存储器(RAM)、静态RAM、动态RAM、只读存储器(ROM)、可编程ROM、可擦除可编程ROM、电可擦除可编程ROM、闪存、安全数字(SD)卡等。
本实施例还提供了一种电子增稳装置，电子增稳装置包括存储器以及处理 器。存储器用于存储可执行指令。处理器用于执行存储器中存储的可执行指令，以执行如下操作：获取图像的当前捕获时刻的成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据，图像为由成像装置捕获；至少部分根据运动信息确定用于增稳图像的目标姿态数据；根据目标姿态数据以及当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据增稳图像。
可以理解地，电子增稳装置可以位于可移动平台，也可以独立于可移动平台并与可移动平台通信连接。当然，电子增稳装置也可以与可移动平台没有任何关联，为独立的电子设备，如集成于成像装置。
运动信息包括移动角速度，处理器具体可以执行如下操作：根据移动角速度与角速度阈值的比较结果，确定用于增稳图像的目标姿态数据。
处理器还可以执行如下操作：获取姿态控制指令，姿态控制指令用于控制成像装置的目标姿态；当移动角速度小于或等于角速度阈值时，确定用于增稳图像的目标姿态数据为与姿态控制指令对应的成像装置的目标姿态。
姿态控制指令可以用于指示通过控制可移动平台的姿态来实现成像装置的目标姿态的控制，成像装置承载于可移动平台，和/或，姿态控制指令用于指示通过控制云台的姿态来实现成像装置的目标姿态的控制，成像装置通过云台承载于可移动平台。
处理器具体可以执行如下操作：当移动角速度大于角速度阈值时，根据图像的捕获时刻的成像装置的测量姿态数据以及预定偏置阈值确定用于增稳图像的目标姿态数据。
角速度阈值可以与图像在处理前后的尺寸差异相关。
角速度阈值可以与图像在处理前后的尺寸差异正相关。
运动信息包括移动角速度，处理器具体可以执行如下操作：获取多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据；根据移动角速度，对多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据进行筛选；根据筛选后的姿态数据，确定用于增稳图像的目标姿态数据。
处理器还可以执行如下操作：根据移动角速度确定低通滤波的截止频率；根据低通滤波的截止频率对多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据进行低通滤波，以得到筛选后的姿态数据。
处理器具体可以执行如下操作：根据移动角速度与移动角速度阈值的比较结果，确定低通滤波的截止频率。
处理器具体可以执行如下操作：当移动角速度小于或等于移动角速度阈值时，确定低通滤波的截止频率为预设截止频率。
处理器具体可以执行如下操作：当移动角速度大于移动角速度阈值时，调整低通滤波的截止频率，以大于预设截止频率。
移动角速度大于移动角速度阈值的幅度越大时，低通滤波的截止频率与预设截止频率的差异幅度可以越大。
多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据可 以包括：当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据。
多个不同时刻中的至少部分位于当前捕获时刻之前；多个不同时刻中的至少另一部分位于当前捕获时刻之后。
多个不同时刻中的该至少部分对应的时长，与多个不同时刻中的该至少另一部分对应的时长相等。
不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据的数量大于2，处理器具体可以执行如下操作：对多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据进行多次子低通滤波；其中，最后一次子低通滤波的截止频率为根据低通滤波的截止频率确定。
除最后一次子低通滤波的截止频率以外的任何一次子低通滤波的截止频率，大于低通滤波的截止频率。
在任一次子低通滤波输出对应的输出数据后，该任一次子低通滤波的输入数据被删除。
至少部分子低通滤波的输入数据为多个不同时刻的成像装置的基于姿态传感器的测量数据获得的姿态数据的部分，至少另一部分子低通滤波的输入数据为其他子低通滤波的输出数据。
该至少另一部分子低通滤波中的所有的子低通滤波的输入数据为该至少部分子低通滤波的输出数据。
处理器还可以执行如下操作：每获取到预设个数的成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的子低通滤波；或每获取到预设时长对应的成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的子低通滤波。
成像装置通过云台承载于可移动平台，云台用于增稳成像装置。
本实施例还提供了一种可移动平台，可移动平台承载有成像装置，成像装置用于捕获图像。可移动平台还包括：存储器，用于存储可执行指令；处理器，用于执行存储器中存储的可执行指令，以执行如下操作：获取图像的当前捕获时刻的成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据，图像为由成像装置捕获；至少部分根据运动信息确定用于增稳图像的目标姿态数据；根据目标姿态数据以及当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据增稳图像。
可以理解，该处理器进一步执行的相关操作可以参照上述电子增稳装置中处理器的执行的相关操作，此处不再赘述。
可移动平台还可以包括云台，可移动平台通过云台承载成像装置，云台用于增稳成像装置。
本实施例还提供了一种成像装置，成像装置包括图像传感器，该图像传感器用于捕获图像。成像装置还包括：存储器，用于存储可执行指令；处理器，用于执行存储器中存储的可执行指令，以执行如下操作：获取图像的当前捕获时刻的成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据，图像为由 成像装置捕获；至少部分根据运动信息确定用于增稳图像的目标姿态数据；根据目标姿态数据以及当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据增稳图像。
可以理解，该处理器进一步执行的相关操作可以参照上述电子增稳装置中处理器的执行的相关操作，此处不再赘述。
可以理解，本实施例中的成像装置可以为诸如数码摄像机、运动相机、单反相机、微单相机等，可以为通过用户手持使用，也可以是通过可移动平台和/或云台使用。
对于本申请的实施例，还需要说明的是，在不冲突的情况下，本申请的实施例及实施例中的特征可以相互组合以得到新的实施例。
以上所述仅为本申请的优选实施例而已，并不用于限制本申请，对于本领域的技术人员来说，本申请可以有各种更改和变化。凡在本申请的精神和原则之内，所作的任何修改、等同替换、改进等，均应包含在本申请的保护范围之内。

Claims (97)

  1. 一种图像的电子增稳方法，其特征在于，包括：
    获取所述图像的当前捕获时刻的成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据，所述图像为由所述成像装置捕获；
    至少部分根据所述运动信息确定用于增稳所述图像的目标姿态数据；
    根据所述目标姿态数据以及所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据对所述图像进行处理。
  2. 根据权利要求1所述的电子增稳方法，其特征在于，所述运动信息包括移动角速度，所述至少部分根据所述运动信息确定用于增稳所述图像的目标姿态数据，包括：
    根据所述移动角速度与角速度阈值的比较结果，确定用于增稳所述图像的目标姿态数据。
  3. 根据权利要求2所述的电子增稳方法，其特征在于，还包括：
    获取姿态控制指令，所述姿态控制指令用于控制所述成像装置的目标姿态；
    所述根据所述移动角速度与角速度阈值的比较结果，确定用于增稳所述图像的目标姿态数据，包括：
    当所述移动角速度小于或等于所述角速度阈值时，确定用于增稳所述图像的目标姿态数据为与所述姿态控制指令对应的所述成像装置的目标姿态。
  4. 根据权利要求3所述的电子增稳方法，其特征在于，所述姿态控制指令用于指示通过控制可移动平台的姿态来实现所述成像装置的目标姿态的控制，所述成像装置承载于所述可移动平台；和/或
    所述姿态控制指令用于指示通过控制云台的姿态来实现所述成像装置的目标姿态的控制，所述成像装置通过所述云台承载于所述可移动平台。
  5. 根据权利要求2所述的电子增稳方法，其特征在于，所述根据所述移动角速度与角速度阈值的比较结果，确定用于增稳所述图像的目标姿态数据，包括：
    当所述移动角速度大于所述角速度阈值时，根据所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据以及预定偏置阈值确定用于增稳所述图像的目标姿态数据。
  6. 根据权利要求2至5中任一项所述的电子增稳方法，其特征在于，所述角速度阈值与所述图像在处理前后的尺寸差异相关。
  7. 根据权利要求6所述的电子增稳方法，其特征在于，所述角速度阈值与所述图像在处理前后的尺寸差异正相关。
  8. 根据权利要求1所述的电子增稳方法，其特征在于，所述运动信息包括移动角速度，还包括：
    获取多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据；
    所述至少部分根据所述运动信息确定用于增稳所述图像的目标姿态数据，包 括：
    根据所述移动角速度，对所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据进行筛选；
    根据筛选后的姿态数据，确定用于增稳所述图像的目标姿态数据。
  9. 根据权利要求8所述的电子增稳方法，其特征在于，所述根据所述移动角速度，对所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据进行筛选，包括：
    根据所述移动角速度确定低通滤波的截止频率；
    根据所述低通滤波的截止频率对所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据进行所述低通滤波，以得到筛选后的姿态数据。
  10. 根据权利要求9所述的电子增稳方法，其特征在于，所述根据所述移动角速度确定低通滤波的截止频率，包括：
    根据所述移动角速度与移动角速度阈值的比较结果，确定所述低通滤波的截止频率。
  11. 根据权利要求10所述的电子增稳方法，其特征在于，所述根据所述移动角速度与移动角速度阈值的比较结果，确定所述低通滤波的截止频率，包括：
    当所述移动角速度小于或等于所述移动角速度阈值时，确定所述低通滤波的截止频率为预设截止频率。
  12. 根据权利要求10所述的电子增稳方法，其特征在于，所述根据所述移动角速度与移动角速度阈值的比较结果，确定所述低通滤波的截止频率，包括：
    当所述移动角速度大于所述移动角速度阈值时，调整所述低通滤波的截止频率，以大于预设截止频率。
  13. 根据权利要求12所述的电子增稳方法，其特征在于，所述移动角速度大于所述移动角速度阈值的幅度越大时，所述低通滤波的截止频率与所述预设截止频率的差异幅度越大。
  14. 根据权利要求8所述的电子增稳方法，其特征在于，所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据包括：所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据。
  15. 根据权利要求14所述的电子增稳方法，其特征在于，所述多个不同时刻中的至少部分位于所述当前捕获时刻之前；
    所述多个不同时刻中的至少另一部分位于所述当前捕获时刻之后。
  16. 根据权利要求15所述的电子增稳方法，其特征在于，所述多个不同时刻中的所述至少部分对应的时长，与所述多个不同时刻中的所述至少另一部分对应的时长相等。
  17. 根据权利要求9所述的电子增稳方法，其特征在于，不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据的数量大于2，所述根据所述低通滤波的截止频率对所述多个不同时刻的所述成像装置的基于姿态传感器 的测量数据获得的姿态数据进行所述低通滤波，包括：
    对所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据进行多次子低通滤波；
    其中，最后一次所述子低通滤波的截止频率为根据所述低通滤波的截止频率确定。
  18. 根据权利要求17所述的电子增稳方法，其特征在于，除最后一次所述子低通滤波的截止频率以外的任何一次所述子低通滤波的截止频率，大于所述低通滤波的截止频率。
  19. 根据权利要求17所述的电子增稳方法，其特征在于，在任一次所述子低通滤波输出对应的输出数据后，所述任一次所述子低通滤波的输入数据被删除。
  20. 根据权利要求17所述的电子增稳方法，其特征在于，至少部分所述子低通滤波的输入数据为所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据的部分，至少另一部分所述子低通滤波的输入数据为其他所述子低通滤波的输出数据。
  21. 根据权利要求20所述的电子增稳方法，其特征在于，所述至少另一部分所述子低通滤波中的所有的所述子低通滤波的输入数据为所述至少部分所述子低通滤波的输出数据。
  22. 根据权利要求17所述的电子增稳方法，其特征在于，所述对所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据进行多次子低通滤波，包括：
    每获取到预设个数的所述成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的所述子低通滤波；或
    每获取到预设时长对应的所述成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的所述子低通滤波。
  23. 根据权利要求1所述的电子增稳方法，其特征在于，所述根据所述目标姿态数据以及所述当前捕获时刻的姿态数据对所述图像进行处理，包括：
    根据所述目标姿态数据以及所述当前捕获时刻的姿态数据对所述图像进行校正。
  24. 根据权利要求23所述的电子增稳方法，其特征在于，所述校正包括旋转校正和/或平移校正。
  25. 一种电子增稳装置，其特征在于，所述电子增稳装置包括：
    存储器，用于存储可执行指令；
    处理器，用于执行所述存储器中存储的所述可执行指令，以执行如下操作：
    获取图像的当前捕获时刻的成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据，所述图像为由所述成像装置捕获；
    至少部分根据所述运动信息确定用于增稳所述图像的目标姿态数据；
    根据所述目标姿态数据以及所述当前捕获时刻的基于姿态传感器的测量数 据获得的姿态数据增稳所述图像。
  26. 根据权利要求25所述的电子增稳装置，其特征在于，所述运动信息包括移动角速度，所述处理器具体执行如下操作：
    根据所述移动角速度与角速度阈值的比较结果，确定用于增稳所述图像的目标姿态数据。
  27. 根据权利要求26所述的电子增稳装置，其特征在于，所述处理器具体执行如下操作：
    获取姿态控制指令，所述姿态控制指令用于控制所述成像装置的目标姿态；
    当所述移动角速度小于或等于所述角速度阈值时，确定用于增稳所述图像的目标姿态数据为与所述姿态控制指令对应的所述成像装置的目标姿态。
  28. 根据权利要求27所述的电子增稳装置，其特征在于，所述姿态控制指令用于指示通过控制可移动平台的姿态来实现所述成像装置的目标姿态的控制，所述成像装置承载于所述可移动平台；和/或
    所述姿态控制指令用于指示通过控制云台的姿态来实现所述成像装置的目标姿态的控制，所述成像装置通过所述云台承载于所述可移动平台。
  29. 根据权利要求26所述的电子增稳装置，其特征在于，所述处理器具体执行如下操作：
    当所述移动角速度大于所述角速度阈值时，根据所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据以及预定偏置阈值确定用于增稳所述图像的目标姿态数据。
  30. 根据权利要求26至29中任一项所述的电子增稳装置，其特征在于，所述角速度阈值与所述图像在处理前后的尺寸差异相关。
  31. 根据权利要求30所述的电子增稳装置，其特征在于，所述角速度阈值与所述图像在处理前后的尺寸差异正相关。
  32. 根据权利要求25所述的电子增稳装置，其特征在于，所述运动信息包括移动角速度，所述处理器还执行如下操作：
    获取多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据；
    根据所述移动角速度，对所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据进行筛选；
    根据筛选后的姿态数据，确定用于增稳所述图像的目标姿态数据。
  33. 根据权利要求32所述的电子增稳装置，其特征在于，所述处理器具体执行如下操作：
    根据所述移动角速度确定低通滤波的截止频率；
    根据所述低通滤波的截止频率对所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据进行所述低通滤波，以得到筛选后的测量姿态数据。
  34. 根据权利要求33所述的电子增稳装置，其特征在于，所述处理器具体 执行如下操作：
    根据所述移动角速度与移动角速度阈值的比较结果，确定所述低通滤波的截止频率。
  35. 根据权利要求34所述的电子增稳装置，其特征在于，所述处理器具体执行如下操作：
    当所述移动角速度小于或等于所述移动角速度阈值时，确定所述低通滤波的截止频率为预设截止频率。
  36. 根据权利要求34所述的电子增稳装置，其特征在于，所述处理器具体执行如下操作：
    当所述移动角速度大于所述移动角速度阈值时，调整所述低通滤波的截止频率，以大于预设截止频率。
  37. 根据权利要求36所述的电子增稳装置，其特征在于，所述移动角速度大于所述移动角速度阈值的幅度越大时，所述低通滤波的截止频率与所述预设截止频率的差异幅度越大。
  38. 根据权利要求32所述的电子增稳装置，其特征在于，所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据包括：所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据。
  39. 根据权利要求38所述的电子增稳装置，其特征在于，所述多个不同时刻中的至少部分位于所述当前捕获时刻之前；
    所述多个不同时刻中的至少另一部分位于所述当前捕获时刻之后。
  40. 根据权利要求39所述的电子增稳装置，其特征在于，所述多个不同时刻中的所述至少部分对应的时长，与所述多个不同时刻中的所述至少另一部分对应的时长相等。
  41. 根据权利要求33所述的电子增稳装置，其特征在于，不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据的数量大于2，所述处理器具体执行如下操作：
    对所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据进行多次子低通滤波；
    其中，最后一次所述子低通滤波的截止频率为根据所述低通滤波的截止频率确定。
  42. 根据权利要求41所述的电子增稳装置，其特征在于，除最后一次所述子低通滤波的截止频率以外的任何一次所述子低通滤波的截止频率，大于所述低通滤波的截止频率。
  43. 根据权利要求41所述的电子增稳装置，其特征在于，在任一次所述子低通滤波输出对应的输出数据后，所述任一次所述子低通滤波的输入数据被删除。
  44. 根据权利要求41所述的电子增稳装置，其特征在于，至少部分所述子低通滤波的输入数据为所述多个不同时刻的所述成像装置的基于姿态传感器的 测量数据获得的姿态数据的部分，至少另一部分所述子低通滤波的输入数据为其他所述子低通滤波的输出数据。
  45. 根据权利要求44所述的电子增稳装置，其特征在于，所述至少另一部分所述子低通滤波中的所有的所述子低通滤波的输入数据为所述至少部分所述子低通滤波的输出数据。
  46. 根据权利要求41所述的电子增稳装置，其特征在于，所述处理器具体执行如下操作：
    每获取到预设个数的所述成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的所述子低通滤波；或
    每获取到预设时长对应的所述成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的所述子低通滤波。
  47. 根据权利要求32所述的电子增稳装置，其特征在于，所述处理器具体执行如下操作：
    根据所述目标姿态数据以及所述当前捕获时刻的姿态数据对所述图像进行校正。
  48. 根据权利要求47所述的电子增稳装置，其特征在于，所述校正包括旋转校正和/或平移校正。
  49. 一种可移动平台，其特征在于，所述可移动平台用于承载成像装置，所述成像装置用于捕获图像；所述可移动平台还包括：
    存储器，用于存储可执行指令；
    处理器，用于执行所述存储器中存储的所述可执行指令，以执行如下操作：
    获取所述图像的当前捕获时刻的所述成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据；
    至少部分根据所述运动信息确定用于增稳所述图像的目标姿态数据；
    根据所述目标姿态数据以及所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据对所述图像进行处理。
  50. 根据权利要求49所述的可移动平台，其特征在于，所述运动信息包括移动角速度，所述处理器具体执行如下操作：
    根据所述移动角速度与角速度阈值的比较结果，确定用于增稳所述图像的目标姿态数据。
  51. 根据权利要求50所述的可移动平台，其特征在于，所述处理器具体执行如下操作：
    获取姿态控制指令，所述姿态控制指令用于控制所述成像装置的目标姿态；
    当所述移动角速度小于或等于所述角速度阈值时，确定用于增稳所述图像的目标姿态数据为与所述姿态控制指令对应的所述成像装置的目标姿态。
  52. 根据权利要求51所述的可移动平台，其特征在于，所述姿态控制指令用于指示通过控制所述可移动平台的姿态来实现所述成像装置的目标姿态的控制；和/或
    所述姿态控制指令用于指示通过控制云台的姿态来实现所述成像装置的目标姿态的控制，所述成像装置通过所述云台承载于所述可移动平台。
  53. 根据权利要求50所述的可移动平台，其特征在于，所述处理器具体执行如下操作：
    当所述移动角速度大于所述角速度阈值时，根据所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据以及预定偏置阈值确定用于增稳所述图像的目标姿态数据。
  54. 根据权利要求50至53中任一项所述的可移动平台，其特征在于，所述角速度阈值与所述图像在处理前后的尺寸差异相关。
  55. 根据权利要求54所述的可移动平台，其特征在于，所述角速度阈值与所述图像在处理前后的尺寸差异正相关。
  56. 根据权利要求49所述的可移动平台，其特征在于，所述运动信息包括移动角速度，所述处理器具体执行如下操作：
    获取多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据；
    根据所述移动角速度，对所述多个不同时刻的所述成像装置的姿态传感器的测量数据获得的姿态数据进行筛选；
    根据筛选后的姿态数据，确定用于增稳所述图像的目标姿态数据。
  57. 根据权利要求56所述的可移动平台，其特征在于，所述处理器具体执行如下操作：
    根据所述移动角速度确定低通滤波的截止频率；
    根据所述低通滤波的截止频率对所述多个不同时刻的基于姿态传感器的测量数据获得的所述成像装置的姿态数据进行所述低通滤波，以得到筛选后的姿态数据。
  58. 根据权利要求57所述的可移动平台，其特征在于，所述处理器具体执行如下操作：
    根据所述移动角速度与移动角速度阈值的比较结果，确定所述低通滤波的截止频率。
  59. 根据权利要求58所述的可移动平台，其特征在于，所述处理器具体执行如下操作：
    当所述移动角速度小于或等于所述移动角速度阈值时，确定所述低通滤波的截止频率为预设截止频率。
  60. 根据权利要求58所述的可移动平台，其特征在于，所述处理器具体执行如下操作：
    当所述移动角速度大于所述移动角速度阈值时，调整所述低通滤波的截止频率，以大于预设截止频率。
  61. 根据权利要求60所述的可移动平台，其特征在于，所述移动角速度大于所述移动角速度阈值的幅度越大时，所述低通滤波的截止频率与所述预设截止 频率的差异幅度越大。
  62. 根据权利要求56所述的可移动平台，其特征在于，所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据包括：所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据。
  63. 根据权利要求62所述的可移动平台，其特征在于，所述多个不同时刻中的至少部分位于所述当前捕获时刻之前；
    所述多个不同时刻中的至少另一部分位于所述当前捕获时刻之后。
  64. 根据权利要求63所述的可移动平台，其特征在于，所述多个不同时刻中的所述至少部分对应的时长，与所述多个不同时刻中的所述至少另一部分对应的时长相等。
  65. 根据权利要求57所述的可移动平台，其特征在于，不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据的数量大于2，所述处理器具体执行如下操作：
    对所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据进行多次子低通滤波；
    其中，最后一次所述子低通滤波的截止频率为根据所述低通滤波的截止频率确定。
  66. 根据权利要求65所述的可移动平台，其特征在于，除最后一次所述子低通滤波的截止频率以外的任何一次所述子低通滤波的截止频率，大于所述低通滤波的截止频率。
  67. 根据权利要求65所述的可移动平台，其特征在于，在任一次所述子低通滤波输出对应的输出数据后，所述任一次所述子低通滤波的输入数据被删除。
  68. 根据权利要求65所述的可移动平台，其特征在于，至少部分所述子低通滤波的输入数据为所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据的部分，至少另一部分所述子低通滤波的输入数据为其他所述子低通滤波的输出数据。
  69. 根据权利要求68所述的可移动平台，其特征在于，所述至少另一部分所述子低通滤波中的所有的所述子低通滤波的输入数据为所述至少部分所述子低通滤波的输出数据。
  70. 根据权利要求65所述的可移动平台，其特征在于，所述处理器具体执行如下操作：
    每获取到预设个数的所述成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的所述子低通滤波；或
    每获取到预设时长对应的所述成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的所述子低通滤波。
  71. 根据权利要求49所述的可移动平台，其特征在于，所述处理器具体执行如下操作：
    根据所述目标姿态数据以及所述当前捕获时刻的姿态数据对所述图像进行 校正。
  72. 根据权利要求71所述的可移动平台，其特征在于，所述校正包括旋转校正和/或平移校正。
  73. 一种成像装置，所述成像装置包括图像传感器，所述图像传感器用于捕获图像，其特征在于，所述成像装置还包括：
    存储器，用于存储可执行指令；
    处理器，用于执行所述存储器中存储的所述可执行指令，以执行如下操作：
    获取所述图像的当前捕获时刻的所述成像装置的运动信息以及基于姿态传感器的测量数据获得的姿态数据；
    至少部分根据所述运动信息确定用于增稳所述图像的目标姿态数据；
    根据所述目标姿态数据以及所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据对所述图像进行处理。
  74. 根据权利要求73所述的成像装置，其特征在于，所述运动信息包括移动角速度，所述处理器具体执行如下操作：
    根据所述移动角速度与角速度阈值的比较结果，确定用于增稳所述图像的目标姿态数据。
  75. 根据权利要求74所述的成像装置，其特征在于，所述处理器具体执行如下操作：
    获取姿态控制指令，所述姿态控制指令用于控制所述成像装置的目标姿态；
    当所述移动角速度小于或等于所述角速度阈值时，确定用于增稳所述图像的目标姿态数据为与所述姿态控制指令对应的所述成像装置的目标姿态。
  76. 根据权利要求75所述的成像装置，其特征在于，所述姿态控制指令用于指示通过控制可移动平台的姿态来实现所述成像装置的目标姿态的控制，所述成像装置承载于所述可移动平台；和/或
    所述姿态控制指令用于指示通过控制云台的姿态来实现所述成像装置的目标姿态的控制，所述成像装置通过所述云台承载于所述可移动平台。
  77. 根据权利要求74所述的成像装置，其特征在于，所述处理器具体执行如下操作：
    当所述移动角速度大于所述角速度阈值时，根据所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据以及预定偏置阈值确定用于增稳所述图像的目标姿态数据。
  78. 根据权利要求74至77中任一项所述的成像装置，其特征在于，所述角速度阈值与所述图像在处理前后的尺寸差异相关。
  79. 根据权利要求78所述的成像装置，其特征在于，所述角速度阈值与所述图像在处理前后的尺寸差异正相关。
  80. 根据权利要求73所述的成像装置，其特征在于，所述运动信息包括移动角速度，所述处理器具体执行如下操作：
    获取多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿 态数据；
    根据所述移动角速度，对所述多个不同时刻的所述成像装置的姿态传感器的测量数据获得的姿态数据进行筛选；
    根据筛选后的姿态数据，确定用于增稳所述图像的目标姿态数据。
  81. 根据权利要求80所述的成像装置，其特征在于，所述处理器具体执行如下操作：
    根据所述移动角速度确定低通滤波的截止频率；
    根据所述低通滤波的截止频率对所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据进行所述低通滤波，以得到筛选后的姿态数据。
  82. 根据权利要求81所述的成像装置，其特征在于，所述处理器具体执行如下操作：
    根据所述移动角速度与移动角速度阈值的比较结果，确定所述低通滤波的截止频率。
  83. 根据权利要求82所述的成像装置，其特征在于，所述处理器具体执行如下操作：
    当所述移动角速度小于或等于所述移动角速度阈值时，确定所述低通滤波的截止频率为预设截止频率。
  84. 根据权利要求82所述的成像装置，其特征在于，所述处理器具体执行如下操作：
    当所述移动角速度大于所述移动角速度阈值时，调整所述低通滤波的截止频率，以大于预设截止频率。
  85. 根据权利要求84所述的成像装置，其特征在于，所述移动角速度大于所述移动角速度阈值的幅度越大时，所述低通滤波的截止频率与所述预设截止频率的差异幅度越大。
  86. 根据权利要求80所述的成像装置，其特征在于，所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据包括：所述当前捕获时刻的基于姿态传感器的测量数据获得的姿态数据。
  87. 根据权利要求86所述的成像装置，其特征在于，所述多个不同时刻中的至少部分位于所述当前捕获时刻之前；
    所述多个不同时刻中的至少另一部分位于所述当前捕获时刻之后。
  88. 根据权利要求87所述的成像装置，其特征在于，所述多个不同时刻中的所述至少部分对应的时长，与所述多个不同时刻中的所述至少另一部分对应的时长相等。
  89. 根据权利要求81所述的成像装置，其特征在于，不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据的数量大于2，所述处理器具体执行如下操作：
    对所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的 姿态数据进行多次子低通滤波；
    其中，最后一次所述子低通滤波的截止频率为根据所述低通滤波的截止频率确定。
  90. 根据权利要求89所述的成像装置，其特征在于，除最后一次所述子低通滤波的截止频率以外的任何一次所述子低通滤波的截止频率，大于所述低通滤波的截止频率。
  91. 根据权利要求89所述的成像装置，其特征在于，在任一次所述子低通滤波输出对应的输出数据后，所述任一次所述子低通滤波的输入数据被删除。
  92. 根据权利要求89所述的成像装置，其特征在于，至少部分所述子低通滤波的输入数据为所述多个不同时刻的所述成像装置的基于姿态传感器的测量数据获得的姿态数据的部分，至少另一部分所述子低通滤波的输入数据为其他所述子低通滤波的输出数据。
  93. 根据权利要求92所述的成像装置，其特征在于，所述至少另一部分所述子低通滤波中的所有的所述子低通滤波的输入数据为所述至少部分所述子低通滤波的输出数据。
  94. 根据权利要求89所述的成像装置，其特征在于，所述处理器具体执行如下操作：
    每获取到预设个数的所述成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的所述子低通滤波；或
    每获取到预设时长对应的所述成像装置的基于姿态传感器的测量数据获得的姿态数据时，进行对应的所述子低通滤波。
  95. 根据权利要求73所述的成像装置，其特征在于，所述处理器具体执行如下操作：
    根据所述目标姿态数据以及所述当前姿态数据对所述图像进行校正。
  96. 根据权利要求95所述的成像装置，其特征在于，所述校正包括旋转校正和/或平移校正。
  97. 一种计算机可读存储介质，其特征在于，其存储有可执行指令，所述可执行指令在由一个或多个处理器执行时，使所述一个或多个处理器执行如权利要求1至24中任一项权利要求所述的电子增稳方法。
PCT/CN2021/077620 2021-02-24 2021-02-24 电子增稳方法及装置、可移动平台、成像装置 WO2022178703A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/077620 WO2022178703A1 (zh) 2021-02-24 2021-02-24 电子增稳方法及装置、可移动平台、成像装置

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
PCT/CN2021/077620 WO2022178703A1 (zh) 2021-02-24 2021-02-24 电子增稳方法及装置、可移动平台、成像装置

Publications (1)

Publication Number Publication Date
WO2022178703A1 true WO2022178703A1 (zh) 2022-09-01

Family

ID=83048627

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/077620 WO2022178703A1 (zh) 2021-02-24 2021-02-24 电子增稳方法及装置、可移动平台、成像装置

Country Status (1)

Country Link
WO (1) WO2022178703A1 (zh)

Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107920861A (zh) * 2015-08-28 2018-04-17 皇家飞利浦有限公司 用于确定运动关系的装置
US20190045127A1 (en) * 2017-08-03 2019-02-07 Canon Kabushiki Kaisha Image pick-up apparatus and control method thereof
WO2019134155A1 (zh) * 2018-01-07 2019-07-11 深圳市大疆创新科技有限公司 图像数据处理方法、设备、平台及存储介质
CN110235431A (zh) * 2019-04-30 2019-09-13 深圳市大疆创新科技有限公司 电子增稳方法、图像采集设备、可移动平台
CN111866395A (zh) * 2020-08-07 2020-10-30 苏州臻迪智能科技有限公司 增稳处理模组、无人机摄像系统以及图像增稳处理方法
CN112166597A (zh) * 2019-08-22 2021-01-01 深圳市大疆创新科技有限公司 图像处理方法、设备和可移动平台

Patent Citations (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107920861A (zh) * 2015-08-28 2018-04-17 皇家飞利浦有限公司 用于确定运动关系的装置
US20190045127A1 (en) * 2017-08-03 2019-02-07 Canon Kabushiki Kaisha Image pick-up apparatus and control method thereof
WO2019134155A1 (zh) * 2018-01-07 2019-07-11 深圳市大疆创新科技有限公司 图像数据处理方法、设备、平台及存储介质
CN110235431A (zh) * 2019-04-30 2019-09-13 深圳市大疆创新科技有限公司 电子增稳方法、图像采集设备、可移动平台
CN112166597A (zh) * 2019-08-22 2021-01-01 深圳市大疆创新科技有限公司 图像处理方法、设备和可移动平台
CN111866395A (zh) * 2020-08-07 2020-10-30 苏州臻迪智能科技有限公司 增稳处理模组、无人机摄像系统以及图像增稳处理方法

Similar Documents

Publication Publication Date Title
US9924100B2 (en) Image-blur correction apparatus, tilt correction apparatus, method of controlling image-blur correction apparatus, and method of controlling tilt correction apparatus
US10958838B2 (en) Method and device for electronic image stabilization of a captured image
US10848676B2 (en) Semiconductor device and electronic device
JP6128389B2 (ja) 撮像装置
JP4518197B2 (ja) 撮像装置、及び像振れ補正方法並びにプログラム
JP6135848B2 (ja) 撮像装置、画像処理装置および画像処理方法
US11025823B2 (en) Electronic stabilization method, image acquisition device, and movable platform
US9131156B2 (en) Amount-of-shake sensing apparatus, imaging apparatus, and method for detecting the amount of shake
CN106210544A (zh) 一种移动终端及其拍摄视频的防抖处理方法、系统
WO2020227998A1 (zh) 图像增稳控制方法、拍摄设备和可移动平台
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
CN111800589A (zh) 图像处理方法、装置和系统，以及机器人
EP3842682A1 (en) Cradle head control method and device, cradle head system and unmanned aerial vehicle
JP2023052772A (ja) 撮像支援装置、撮像装置、撮像システム、撮像支援システム、撮像支援方法、及びプログラム
JP2019145958A (ja) 撮像装置およびその制御方法ならびにプログラム
EP4221181A1 (en) Method for generating rotation direction of gyroscope and computer device
WO2022178703A1 (zh) 电子增稳方法及装置、可移动平台、成像装置
WO2022067548A1 (zh) 云台的控制方法及装置、云台以及可移动平台
JP7324284B2 (ja) 撮像デバイス、手ぶれ補正デバイス、撮像方法及び手ぶれ補正方法
CN110832424A (zh) 竖向增稳机构及其控制方法以及可移动设备
WO2022134036A1 (zh) 可移动平台及其控制方法、装置
US20190260932A1 (en) Image stabilization control device for imaging device
WO2022061915A1 (zh) 电机及其控制方法、计算机可读存储介质以及机械装置
WO2021106454A1 (ja) 撮像支援装置、撮像装置、撮像システム、撮像支援方法、及びプログラム
US9270884B2 (en) Imaging apparatus and detecting apparatus

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 21927148

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 21927148

Country of ref document: EP

Kind code of ref document: A1