CN106488081B - Video Stabilization system and method - Google Patents
Video Stabilization system and method Download PDFInfo
- Publication number
- CN106488081B CN106488081B CN201610903521.7A CN201610903521A CN106488081B CN 106488081 B CN106488081 B CN 106488081B CN 201610903521 A CN201610903521 A CN 201610903521A CN 106488081 B CN106488081 B CN 106488081B
- Authority
- CN
- China
- Prior art keywords
- video
- module
- data
- camera unit
- attitudes vibration
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Active
Links
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N5/00—Details of television systems
- H04N5/14—Picture signal circuitry for video frequency region
- H04N5/21—Circuitry for suppressing or minimising disturbance, e.g. moiré or halo
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04N—PICTORIAL COMMUNICATION, e.g. TELEVISION
- H04N23/00—Cameras or camera modules comprising electronic image sensors; Control thereof
- H04N23/60—Control of cameras or camera modules
- H04N23/68—Control of cameras or camera modules for stable pick-up of the scene, e.g. compensating for camera body vibrations
- H04N23/681—Motion detection
- H04N23/6812—Motion detection based on additional sensors, e.g. acceleration sensors
Landscapes
- Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Signal Processing (AREA)
- Studio Devices (AREA)
Abstract
The present invention relates to a kind of Video Stabilization system and method, existing surely as processing system is unable to reach high frame-rate video real time handling requirement to solve the problems, such as, this system includes: the camera unit for acquiring video data；For obtaining the camera unit attitudes vibration and output phase answers the posture acquiring units of 3 d pose data；It is electrically connected with the posture acquiring unit, the compensation calculation unit for the transformation matrix according to needed for calculating the 3 d pose data and export the compensation camera unit attitudes vibration；It is electrically connected with the compensation calculation unit and the camera unit, for carrying out the video processing unit of attitudes vibration elimination to the video data that the camera unit exports according to the transformation matrix.The image stabilization system can realize that output 1080P resolution ratio, the steady picture effect of 40 frame per second, video output delay are lower than 100ms, can easily play out viewing using the way of output of RTSP video flowing, greatly facilitate commercialization in real time.
Description
Technical field
The present invention relates to technical field of image processing, more specifically to a kind of Video Stabilization system and method.
Background technique
For the capture apparatus of free-standing, how to shoot stable image is research emphasis and difficulty all the time
Point.The Key technique problem of steady picture generally comprises: the acquisition of motion vector and the realization of Contrary compensation.The acquisition side of motion vector
Formula specifically includes that images match and aiding sensors.The implementation of Contrary compensation specifically includes that mechanical compensation and image compensation
Two ways.Wherein mechanical compensation needs that aiding sensors is cooperated to use, and image compensation can then be obtained with cooperative movement vector
Mode is realized.It is steady as solution mainly includes the following three types based on above-mentioned technology:
1, optical anti-vibration: reaching steady as effect by compensating to imaging optical path, including eyeglass movable type and photosensitive
Element is mobile.
2, the steady picture of holder: by the way that video camera is mounted on the mechanical steady holder of increasing, video camera is physically eliminated
Shake, and then prevent the generation of video jitter.
3, electronic steady image: the video that video camera is filmed is post-processed to eliminate video jitter by computer.
But above-mentioned three kinds of schemes have certain limitation:
1, optical anti-vibration needs to install gyroscope in the camera to obtain camera shake, then passes through high-accuracy mechanical structure
Mobile eyeglass or photosensitive element compensate, and guarantee the stabilization of optical path.Mechanical structure used in this mode needs very high work
Skill requirement, it is expensive.
2, for holder surely as required high-performance mechanical cradle head structure is more complicated, product cost is higher；Furthermore holder
Volume is generally all bigger, there is certain use environment limitation.
3, existing electronic steady image is since the complexity and precision problem and DE Camera Shake of vision algorithm are to picture exposure
The influence of quality, it is steady as effect and performance are lower than optical anti-vibration and the steady picture of holder.
Chinese patent application (CN201510283660X) discloses a kind of electronic image stabilization method of mobile terminal video, leads to
The exercise data for being built in the attitude transducer acquisition mobile terminal of mobile terminal is crossed, and by CPU built in mobile terminal to every
One frame moving image carries out translation or rotation processing realizes steady picture.But in actually calculating, CPU is needed on every frame moving image
Each pixel traversed, greatly reduce processing speed.Image is bigger, and its processing speed is slower, the processing system and side
Method is unable to reach the real time handling requirement to high frame-rate video.
Summary of the invention
The technical problem to be solved in the present invention is that cannot achieve the defect of the steady picture of high performance video for the prior art,
A kind of Video Stabilization system and method is provided.
The technical solution adopted by the present invention to solve the technical problems is: constructing a kind of Video Stabilization system: including:
For acquiring the camera unit of video data；
For obtaining the camera unit attitudes vibration and output phase answers the posture acquiring units of 3 d pose data；
Be electrically connected with the posture acquiring unit, for according to the 3 d pose data calculate and export compensation described in take the photograph
The compensation calculation unit of the transformation matrix as needed for unit attitudes vibration；
It is electrically connected, is used for according to the transformation matrix to the camera shooting with the compensation calculation unit and the camera unit
The video data of unit output carries out the video processing unit of attitudes vibration elimination.
Preferably, the compensation calculation unit includes:
The acquisition module for obtaining 3 d pose data is connect with the posture acquiring unit；
It is calculated to the acquisition module connection for calculating the matrix of corresponding transformation matrix according to the 3 d pose data
Module；
It is connect with the matrix computing module and the video processing unit, for exporting the transformation matrix to institute
State the Output matrix module of video processing unit.
Preferably, the video processing unit includes:
The configuration module for carrying out initial configuration to above-mentioned camera unit is connect with the camera unit；
The video acquisition module for receiving the camera unit output video data is connect with the camera unit；
It is connect with the video acquisition module, the video received video data being buffered in internal FIFO
Cache module；
For storing the memory module of video data；
It is connect with the video cache module, memory module, for being written and read the control of control to the memory module
Module；
It connect, is used for according to the transformation matrix to video data with the control module and the Output matrix module
Affine transformation is carried out to eliminate the attitudes vibration cancellation module of attitudes vibration, the attitudes vibration cancellation module passes through the control
Module will carry out affine transformation, and treated that video data is exported to memory module.
Preferably, the video acquisition module connect with the acquisition module and sends down trigger to the acquisition module
Signal；The acquisition module exports 3 d pose data to the matrix computing module according to the interrupt trigger signal.
Preferably, the video processing unit further include:
It connect with the control module, is read by the video that the control module reads video data in the memory module
Modulus block；
It is connect with the video read module, for generating normal video output timing and being exported according to the normal video
The timing generation module of timing reading video data from the video read module；
It is connect with the timing generation module, for exporting the Video Output Modules of video data.
Preferably, the compensation calculation unit further include:
It is connect with the Video Output Modules, for receiving the video reception module of video data；
It is connect with the video reception module, for carrying out the compressed encoding mould of compressed encoding processing to the video data
Block；
It is connect with the compressed encoding module, for the video data after progress compressed encoding to be packaged and pass through
RTSP agreement is exported to the network video stream generation module of external equipment.
Implement image stabilization system of the invention, can realize high performance view under conditions of not increasing complicated external hardware
Frequency can accomplish and common angle attitudes vibration modified identical effect the amendment of wide-angle attitudes vibration surely as effect
Fruit.The image stabilization system can realize that output 1080P resolution ratio, the steady picture effect of 40 frame per second, video output delay are lower than in real time
100ms can easily play out viewing using the way of output of RTSP video flowing, greatly facilitate commercialization.
The present invention also provides a kind of video image stabilization methods, comprising the following steps:
S1: the attitudes vibration for acquiring camera unit obtains corresponding 3 d pose data；
S2: according to transformation matrix needed for the 3 d pose compensation data calculation camera unit attitudes vibration；
S3: attitudes vibration Processing for removing is carried out to video data according to the transformation matrix.
Preferably, the 3 d pose data include Eulerian angles (α, β, θ).
Preferably, the step S2 the following steps are included:
S21: the field angle V of the camera unit is obtained；
S22: the horizontal resolution W and vertical resolution H of the camera unit are obtained；
S23: transformation matrix H is calculated according to the Eulerian angles, field angle V, horizontal resolution W and vertical resolution H:
Preferably, the step S3 includes:
S31: the original coordinates (x, y) of each pixel are obtained；
S32: to above-mentioned original coordinates carry out affine transformation to eliminate attitudes vibration, obtain pixel new coordinate (x ',
Y '):
Digital image stabilization method of the invention obtains the attitudes vibration parameter of camera, and the matrix for eliminating attitudes vibration is calculated,
And video data is handled according to transformation matrix.The calculating process of this method is simple, as a result reliable, it can be achieved that any angle
Attitudes vibration eliminate, including the complete amendment rotated to 360 ° of full angles of image.
Detailed description of the invention
Present invention will be further explained below with reference to the attached drawings and examples, in attached drawing:
Fig. 1 is the schematic diagram of image stabilization system of the present invention；
Fig. 2 is the structural schematic diagram of image stabilization system shown in FIG. 1；
Fig. 3 is the coordinate system and camera unit lens direction schematic diagram of posture acquiring unit shown in Fig. 1；
Fig. 4 is the flow chart of digital image stabilization method of the present invention.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to the accompanying drawings and embodiments, right
The present invention is further elaborated.
As shown in Figure 1, the present invention provides a kind of Video Stabilization system: including: the camera unit for acquiring video data
1；For obtaining 1 attitudes vibration of camera unit and output phase answers the posture acquiring units 2 of 3 d pose data；It is obtained with posture single
Member 2 is electrically connected, the benefit for the transformation matrix according to needed for calculating 3 d pose data and export compensation 1 attitudes vibration of camera unit
Repay computing unit 3；It is electrically connected with compensation calculation unit 3 and camera unit 1, for being exported according to transformation matrix to camera unit 1
Video data carry out attitudes vibration elimination video processing unit 4.
Camera unit 1 for acquiring video is by camera lens, the camera of the components such as cmos sensor composition, and the present invention adopts
Camera model OV5658 can be 2592*1944, frame per second with acquisition resolution by the configuration to cmos sensor
For the video data of 30fps, data-interface is LVDS interface；Certainly, sensor, the model of camera, type, can basis
Actual needs is selected.
To realize high performance 3 d pose Orientation differences collection effect, posture acquiring unit 2 of the invention uses posture
Sensor obtains the real-time attitude of camera unit 1, the model MPU9250 of the attitude transducer in the present embodiment, inside
It is integrated with three-axis gyroscope, 3-axis acceleration and with data movement processor etc., certainly, the model of attitude transducer can also root
It is adjusted according to actual needs.By configuration data movement processor, the real-time three-dimensional posture of diversified forms can be directly exported
Data, the 3 d pose data that the present invention configures are quaternary number.The configuration of its data output interface is flexible, comprising: I2C, SPI and
Serial ports.The present invention exports 3 d pose data by SPI protocol.
As shown in figure 3, the coordinate system of attitude transducer itself includes X-axis, Y-axis and Z axis, wherein the positive direction of X-axis with
Towards identical, attitude transducer is arranged in parallel the camera lens of camera unit 1 with camera unit 1, to guarantee that attitude transducer can be quasi-
Really capture the attitudes vibration of camera unit 1.
For the real time handling requirement for realizing the high frame-rate video of high-resolution, compensation calculation unit 3 of the invention uses ARM
Plate, wherein Hi3516A is thought in the model sea of ARM chip；Certainly, compensation calculation unit 3 can also use the control of other frameworks
Plate, chip model also can according to need and selected.Video processing unit 4 uses FPGA plate, and the model of fpga chip is matched
Sentos XC7A75T；Certainly, chip model also can according to need and be selected.Camera, attitude transducer, FPGA plate and ARM
It is connected between plate by ffc soft arranging wire (flexible flat cable Flexible Flat Cable)；It should be understood that each device it
Between connection can also be attached by pcb board or other interfaces.Wherein, the data between attitude transducer and ARM plate
Transport protocol is SPI protocol.Signal transmission between camera and FPGA plate includes one group of LVDS signal and one group of IIC signal.
Signal transmission between FPGA plate and ARM plate will embody in the connection relationship of each module below.It should be understood that ARM chip
With fpga chip alternatively at other models with similar functions, ARM chip can also be replaced with to DSP or MCU, by FPGA core
Piece replaces with CPLD.
As shown in Fig. 2, including: the acquisition mould being connect with posture acquiring unit 2 for obtaining 3 d pose data on ARM plate
Block 31；The matrix computing module 32 for calculating corresponding transformation matrix according to 3 d pose data is connected to module 31 is obtained；With
Matrix computing module 32 and video processing unit 4 connect, defeated for transformation matrix to be exported the matrix to video processing unit 4
Module 33 out.
It obtains module 31 to communicate with attitude transducer with SPI protocol, obtains module 31 as SPI host and constantly read conduct
Quaternary number in the attitude transducer of slave is simultaneously converted into Eulerian angles (α, β, θ) output to matrix computing module 32, wherein α is
The angle rotated about the z axis, β are the angle rotated around Y-axis, and θ is the angle rotated around X-axis.Matrix computing module 32 is according to Euler
Angle calculates transformation matrix H:
Wherein, V is the field angle of camera, and W and H are respectively image level and vertical resolution.The specification of camera is joined
Its field angle and its resolution ratio for acquiring image can all be write in number exactly.It is previously written selected by image stabilization system and takes the photograph in ARM
As head specifications parameter, matrix computing module 32 can call directly the parameter and be calculated.Output matrix module 33 and FPGA phase
Even, and with SPI protocol transformation matrix is sent.It should be understood that the calculating process that quaternary number is converted into Eulerian angles can be by matrix
Computing module executes, and is only used as obtaining quaternary number from attitude transducer and send to matrix computing module at this point, obtaining module
Quaternary number.
As shown in Fig. 2, including: to be connect with camera unit 1 for carrying out initial configuration to above-mentioned camera unit 1 on FPGA plate
Configuration module 41；The video acquisition module 42 that video data is exported for receiving camera unit 1 is connect with camera unit 1；With
Video acquisition module 42 connects, the video cache module 43 received video data being buffered in internal FIFO；For
Store the memory module 46 of video data；It is connect with video cache module 43, memory module 46, for being carried out to memory module 46
The control module 44 of Read-write Catrol；It connect, is used for according to transformation matrix to view with control module 44 and Output matrix module 33
For frequency according to affine transformation is carried out to eliminate the attitudes vibration cancellation module 45 of attitudes vibration, attitudes vibration cancellation module 45 passes through control
Molding block 44 will carry out affine transformation, and treated that video data is exported to memory module 46.
Further, video acquisition module 42 connect with acquisition module 31 and sends down trigger letter to module 31 is obtained
Number；It obtains module 31 and 3 d pose data is exported to matrix computing module 32 according to interrupt trigger signal.
The routine configuration that is configured to cmos image sensor of the configuration module 41 to camera unit 1, comprising: initialization is matched
Set, pll clock configuration, interface configuration, resolution ratio configuration, frame per second configuration, exposure configuration, gain configuration etc..What the present invention configured
Interface is IIC interface, is 2592*1944, frame per second 30fps with the resolution ratio postponed.It should be understood that according to reality output need
It wants, different parameter configurations can be carried out, such as resolution ratio is changed to 1920*1080.
Camera unit 1 shoots video according to configuration parameter and sends video counts to video acquisition module 42 with LVDS agreement
According to.LVDS signal is converted into parallel video data and is then passed to video cache module 43 by video acquisition module 42.To protect
The Eulerian angles that obtain can be accurately corresponding with video frame rate in card ARM plate, video acquisition module 42 according to video acquisition frame rate, to
The acquisition module 31 of ARM plate sends GPIO interrupt trigger signal.Whenever FPGA sends a GPIO interrupt trigger signal to acquisition
Module 31 obtains module 31 and just sends primary corresponding Eulerian angles to matrix computing module 32.
Control module 44 except outer for being written and read control to memory module 46, also as each module in entire FPGA plate and
The bridge of the exchange data of memory module 46.Memory module 46 can select DDR SDRAM or other memories.
Video cache module 43 by parallel video data buffer storage in internal FIFO after, by AXI4 agreement by video counts
Burst transmission mode is sent to control module 44 accordingly, and video data is sent to attitudes vibration cancellation module by control module 44
45.Attitudes vibration cancellation module 45 based on the received on every frame video image each pixel original coordinates (x, y) and ARM
The transformation matrix H of output calculates the new coordinate (x ', y ') of each pixel to eliminate attitudes vibration, and calculation method is as follows:
After the new coordinate of each pixel is calculated, so that it may each frame video image for the attitudes vibration that has been eliminated.
Treated the video data of attitudes vibration cancellation module 45 is stored by the output of control module 44 into DDR SDRAM.
Further, image stabilization system further includes the output par, c of video.Specifically, video processing unit 4 further include: with control
Molding block 44 connects, and the video read module 47 of video data in memory module 46 is read by control module 44；It is read with video
Modulus block 47 connects, for generating normal video output timing and according to normal video output timing from video read module 47
The timing generation module 48 of reading video data；It is connect with timing generation module 48, the video for exporting video data exports
Module 49.
Compensation calculation unit 3 further include: connect with Video Output Modules 49, for receiving the video reception mould of video data
Block 34；It is connect with video reception module 34, for carrying out the compressed encoding module 35 of compressed encoding processing to video data；With pressure
Contracting coding module 35 connects, and exports for the video data after progress compressed encoding to be packaged and pass through RTSP agreement to outer
The network video stream generation module 36 of portion's equipment.
Video read module 47 is stored in DDR SDRAM with burst transmission mode by the reading of control module 44
It is buffered in internal FIFO by the video data of attitudes vibration Processing for removing, and by the video data.Timing generation module
48 from the FIFO reading video data of video read module 47 and export according to the video output timing of standard and export mould to video
Block 49, Video Output Modules 49 send the parallel video data and clock signal that receive to the video reception mould on ARM plate
Parallel video data are carried out compressed encoding by block 34, compressed encoding module 35, and compression standard can be for H.264 or H.265.Network
Video flowing generation module 36 exports the video after compressed encoding to external equipment by network mode, and the video flowing way of output is
RTSP.It should be understood that compress mode and outputting standard can be substituted for other modes and standard according to the actual situation, such as will
Video compress mode is changed to MPEG compression, and RTSP is changed to privately owned transport protocol etc..
It should be understood that in some embodiments, the information obtained from attitude transducer can be original acceleration information,
Attitude Calculation is carried out to the original acceleration information again on ARM or other processors, finally obtains transformation matrix.
In some embodiments, camera unit 1, posture acquiring unit 2, compensation calculation unit 3 and video processing unit 4 can
It is fixedly connected into the steady as mould group of a rigidity, is convenient for commercialization.
Implement image stabilization system of the invention, can realize high performance view under conditions of not increasing complicated external hardware
Frequently surely as effect, which can realize that output 1080P resolution ratio, the steady picture effect of 40 frame per second, video output are prolonged in real time
It is lower than 100ms late, viewing can be easily played out using the way of output of RTSP video flowing, greatly facilitate product
Change.
According to above system, the present invention also provides a kind of video image stabilization method, include the following steps (as shown in Figure 4):
S1: the attitudes vibration for acquiring camera unit obtains corresponding 3 d pose data；
S2: according to transformation matrix needed for 3 d pose compensation data calculation camera unit attitudes vibration；
S3: attitudes vibration Processing for removing is carried out to video data according to transformation matrix.
As shown in figure 3, by the way that attitude transducer is arranged in parallel on camera unit to guarantee that attitude transducer can be accurate
The attitudes vibration of camera unit is captured, further, 3 d pose data include Eulerian angles (α, β, θ).Attitude transducer itself
Coordinate system include X-axis, Y-axis and Z axis, wherein the positive direction of X-axis and the camera lens of camera unit are towards identical.In Eulerian angles
α is the angle rotated about the z axis, and β is the angle rotated around Y-axis, and θ is the angle rotated around X-axis.
Further, step S2 the following steps are included:
S21: the field angle V of camera unit is obtained；
S22: the horizontal resolution W and vertical resolution H of camera unit are obtained；
S23: transformation matrix H is calculated according to Eulerian angles, field angle V, horizontal resolution W and vertical resolution H:
Its field angle and its resolution ratio for acquiring image can all be write in the specifications parameter of camera exactly.In calculation procedure
It is previously written surely as selected camera specifications parameter, the parameter can be called directly and calculated.
Further, step S3 includes:
S31: the original coordinates (x, y) of each pixel are obtained；
S32: to above-mentioned original coordinates carry out affine transformation to eliminate attitudes vibration, obtain pixel new coordinate (x ',
Y '):
After the new coordinate of each pixel is calculated, so that it may each frame video image for the attitudes vibration that has been eliminated.
It should be understood that in some embodiments, the information obtained from attitude transducer can be original acceleration information, in subsequent meter
Attitude Calculation is carried out to the original acceleration information again in calculation, finally obtains transformation matrix.
Elimination attitudes vibration is calculated on the basis of obtaining camera attitudes vibration parameter in digital image stabilization method of the invention
Matrix, and video data is handled according to transformation matrix.The calculating process of this method is simple, it is as a result reliable, it can be achieved that
Attitudes vibration is eliminated at any angle, including the complete amendment rotated to 360 ° of full angles of image.
It should be understood that above embodiments only express the preferred embodiment of the present invention, description is more specific and detailed
Carefully, but it cannot be understood as limitations on the scope of the patent of the present invention；It should be pointed out that for the common skill of this field
For art personnel, without departing from the inventive concept of the premise, above-mentioned technical characterstic can be freely combined, can also be done
Several modifications and improvements out, these are all within the scope of protection of the present invention；Therefore, all to be done with scope of the invention as claimed
Equivalents and modification, should belong to the covering scope of the claims in the present invention.
Claims (5)
1. a kind of Video Stabilization system, which is characterized in that including
For acquiring the camera unit (1) of video data；
For obtaining the camera unit (1) attitudes vibration and output phase answers the posture acquiring units (2) of 3 d pose data；
Be electrically connected with the posture acquiring unit (2), for according to the 3 d pose data calculate and export compensation described in take the photograph
The compensation calculation unit (3) of the transformation matrix as needed for unit (1) attitudes vibration；
It is electrically connected with the compensation calculation unit (3) and the camera unit (1), for being taken the photograph according to the transformation matrix to described
As the video data of unit (1) output carries out the video processing unit (4) of attitudes vibration elimination；
The compensation calculation unit (3) includes:
The acquisition module (31) for obtaining 3 d pose data is connect with the posture acquiring unit (2)；
It is calculated to acquisition module (31) connection for calculating the matrix of corresponding transformation matrix according to the 3 d pose data
Module (32)；
It is connect with the matrix computing module (32) and the video processing unit (4), for exporting the transformation matrix
To the Output matrix module (33) of the video processing unit (4)；
The matrix computing module (32) is according to Eulerian angles (α, β, θ), field angle V, horizontal resolution W and vertical resolution H
Calculate transformation matrix H:
;
The video processing unit (4) includes:
The configuration module (41) for carrying out initial configuration to above-mentioned camera unit (1) is connect with the camera unit (1)；
The video acquisition module for receiving the camera unit (1) output video data is connect with the camera unit (1)
(42)；
It is connect with the video acquisition module (42), the video received video data being buffered in internal FIFO
Cache module (43)；
For storing the memory module (46) of video data；
It is connect with the video cache module (43), memory module (46), for being written and read control to the memory module (46)
The control module (44) of system；
It connect, is used for according to the transformation matrix to video with the control module (44) and the Output matrix module (33)
Data carry out affine transformation to eliminate the attitudes vibration cancellation module (45) of attitudes vibration, the attitudes vibration cancellation module (45)
To carry out affine transformation by the control module (44), treated that video data is exported to memory module (46)；
The attitudes vibration cancellation module (45) is used for according to the original coordinates (x, y) and the transformation matrix for receiving pixel
H, and corresponding new coordinate (x ', y ') is calculated according to the following formula:
。
2. Video Stabilization system according to claim 1, which is characterized in that the video acquisition module (42) is obtained with described
Modulus block (31) connection simultaneously sends interrupt trigger signal to the acquisition module (31)；The acquisition module (31) is according in described
Disconnected trigger signal exports 3 d pose data to the matrix computing module (32).
3. Video Stabilization system according to claim 1, which is characterized in that the video processing unit (4) further include:
It is connect with the control module (44), reads video data in the memory module (46) by the control module (44)
Video read module (47)；
It is connect with the video read module (47), for generating normal video output timing and being exported according to the normal video
The timing generation module (48) of timing reading video data from the video read module (47)；
It is connect with the timing generation module (48), for exporting the Video Output Modules (49) of video data.
4. Video Stabilization system according to claim 2, which is characterized in that the compensation calculation unit (3) further include:
It is connect with the Video Output Modules (49), for receiving the video reception module (34) of video data；
It is connect with the video reception module (34), for carrying out the compressed encoding mould of compressed encoding processing to the video data
Block (35)；
It is connect with the compressed encoding module (35), for the video data after progress compressed encoding to be packaged and pass through
RTSP agreement is exported to the network video stream generation module (36) of external equipment.
5. a kind of video image stabilization method, which comprises the following steps:
S1: the attitudes vibration for acquiring camera unit obtains corresponding 3 d pose data；
S2: according to transformation matrix needed for the 3 d pose compensation data calculation camera unit attitudes vibration；
S3: attitudes vibration Processing for removing is carried out to video data according to the transformation matrix；
The 3 d pose data include Eulerian angles (α, β, θ)；
The step S2 the following steps are included:
S21: the field angle V of the camera unit is obtained；
S22: the horizontal resolution W and vertical resolution H of the camera unit are obtained；
S23: transformation matrix is calculated according to the Eulerian angles (α, β, θ), field angle V, horizontal resolution W and vertical resolution H
H:
；
The step S3 includes:
S31: the original coordinates (x, y) of each pixel are obtained；
S32: affine transformation is carried out to eliminate attitudes vibration to above-mentioned original coordinates, obtains the new coordinate (x ', y ') of pixel:
。
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610903521.7A CN106488081B (en) | 2016-10-17 | 2016-10-17 | Video Stabilization system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610903521.7A CN106488081B (en) | 2016-10-17 | 2016-10-17 | Video Stabilization system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106488081A CN106488081A (en) | 2017-03-08 |
| CN106488081B true CN106488081B (en) | 2019-06-28 |
Family
ID=58269732
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610903521.7A Active CN106488081B (en) | 2016-10-17 | 2016-10-17 | Video Stabilization system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106488081B (en) |
Families Citing this family (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN108307119A (en) * | 2018-02-10 | 2018-07-20 | 安徽工程大学 | A kind of moving camera attitude-control device simplified |
| WO2019178872A1 (en) | 2018-03-23 | 2019-09-26 | 华为技术有限公司 | Video image anti-shake method and terminal |
| CN110674665B (en) * | 2018-07-03 | 2023-06-30 | 杭州海康威视系统技术有限公司 | Image processing method and device, forest fire prevention system and electronic equipment |
| CN109462746B (en) * | 2018-10-26 | 2020-11-06 | 北京双髻鲨科技有限公司 | Image stabilization method and device |
| WO2020258164A1 (en) * | 2019-06-27 | 2020-12-30 | 深圳市大疆创新科技有限公司 | Target tracking method and device, and computer storage medium |
| CN111212224A (en) * | 2020-01-10 | 2020-05-29 | 上海摩象网络科技有限公司 | Anti-shake processing method and device applied to image shooting equipment and electronic equipment |
| CN112733773A (en) * | 2021-01-18 | 2021-04-30 | 上海商汤智能科技有限公司 | Object detection method and device, computer equipment and storage medium |
| CN113766132A (en) * | 2021-09-16 | 2021-12-07 | 武汉虎客影像文化传播有限公司 | Video shooting method and device |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1719329A (en) * | 2004-07-09 | 2006-01-11 | 宾得株式会社 | Camera provided with camera-shake compensation functionality |
| CN103763483A (en) * | 2014-01-23 | 2014-04-30 | 贝壳网际（北京）安全技术有限公司 | Method and device for shaking resistance in photo taking process of mobile terminal and mobile terminal |
| CN103841297A (en) * | 2012-11-23 | 2014-06-04 | 中国航天科工集团第三研究院第八三五七研究所 | Electronic image-stabilizing method suitable for resultant-motion camera shooting carrier |
| CN105472373A (en) * | 2015-11-18 | 2016-04-06 | 中国兵器工业计算机应用技术研究所 | Bionic electronic image stabilizing method based on vestibular reflection mechanism and device |
| CN105681674A (en) * | 2016-03-21 | 2016-06-15 | 南通大学 | Image stabilizing method and compound image stabilizing system based on mechanical image stabilizing and electronic image stabilizing |
Family Cites Families (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| GB2474886A (en) * | 2009-10-30 | 2011-05-04 | St Microelectronics | Image stabilisation using motion vectors and a gyroscope |
-
2016
- 2016-10-17 CN CN201610903521.7A patent/CN106488081B/en active Active
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1719329A (en) * | 2004-07-09 | 2006-01-11 | 宾得株式会社 | Camera provided with camera-shake compensation functionality |
| CN103841297A (en) * | 2012-11-23 | 2014-06-04 | 中国航天科工集团第三研究院第八三五七研究所 | Electronic image-stabilizing method suitable for resultant-motion camera shooting carrier |
| CN103763483A (en) * | 2014-01-23 | 2014-04-30 | 贝壳网际（北京）安全技术有限公司 | Method and device for shaking resistance in photo taking process of mobile terminal and mobile terminal |
| CN105472373A (en) * | 2015-11-18 | 2016-04-06 | 中国兵器工业计算机应用技术研究所 | Bionic electronic image stabilizing method based on vestibular reflection mechanism and device |
| CN105681674A (en) * | 2016-03-21 | 2016-06-15 | 南通大学 | Image stabilizing method and compound image stabilizing system based on mechanical image stabilizing and electronic image stabilizing |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106488081A (en) | 2017-03-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106488081B (en) | Video Stabilization system and method | |
| JP6702323B2 (en) | Camera module, solid-state imaging device, electronic device, and imaging method | |
| US20170085740A1 (en) | Systems and methods for storing images and sensor data | |
| CN106210544B (en) | Mobile terminal and anti-shake processing method and system for video shot by mobile terminal | |
| US10897573B2 (en) | Image capturing system, terminal and computer readable medium which correct images | |
| JP2013214952A (en) | Imaging device | |
| CN110035228A (en) | Camera stabilization system, method, electronic equipment and computer readable storage medium | |
| US10506163B2 (en) | Systems and methods for synchronizing sensor data | |
| CN107113376A (en) | A kind of image processing method, device and video camera | |
| WO2022121260A1 (en) | Projection picture anti-shake method and apparatus, projection device, and storage medium | |
| CN106791360A (en) | Generate the method and device of panoramic video | |
| JP2017017689A (en) | Imaging system and program of entire-celestial-sphere moving image | |
| CN109194875A (en) | A kind of image pickup method and electronic equipment | |
| CN105744132B (en) | Optical lens accessory for panoramic image shooting | |
| CN104394316B (en) | Camera pan-tilt posture testing system | |
| CN114339102A (en) | Video recording method and device | |
| JP7031280B2 (en) | Image processing equipment, image processing systems and programs | |
| WO2023005355A1 (en) | Image anti-shake method and electronic device | |
| WO2018024239A1 (en) | Hybrid image stabilization system | |
| JP2020167600A (en) | Image processing method, image processing apparatus, and head-mounted display | |
| WO2018019039A1 (en) | Image synthesis method and device for mobile terminal, and mobile terminal | |
| JP7025042B2 (en) | Spherical image generation method, spherical image generation and display method, spherical image generation system program, and spherical image generation and display system program | |
| CN112752086B (en) | Image signal processor, method and system for environment mapping | |
| CN113489878A (en) | Electronic device, information synchronization method, and computer-readable storage medium | |
| CN115278086B (en) | Electronic anti-shake method for gyroscope |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |