KR100524743B1 - A device of correcting shake recording for video camera - Google Patents

Abstract

The present invention relates to a device for compensating for shaking by extracting a motion vector of a frame-based video signal photographed by a video camera and reducing the size of the movement direction. A discrete time conversion unit for outputting; A quantization unit for quantizing the video signal applied from the discrete time conversion unit by a control signal; A compression unit for outputting a signal of a quantum unit as a video signal compressed by a variable length conversion; An inverse quantum unit for inversely quantizing a signal of the quantum unit; An inverse discrete time conversion unit for converting the frequency signal of the inverse quantum unit into inverse discrete time and outputting the digital signal; A motion estimation unit for estimating a motion vector by comparing the inverse discrete time conversion unit signal with the input image signal; A compensator for calculating and outputting a value for compensating for the shaking of the video signal applied from the inverse discrete time conversion unit based on the motion vector value of the motion estimation unit; The control and monitoring of the quantum conversion of the quantum part, and the control unit for outputting the inter-operation mode control signal for compensating for the motion vector, and the intra-operation mode control signal for compensating for the motion vector, the motion or vibration effect Minimize the shaking caused by the effect is to remove.

Description

Image stabilization device {A DEVICE OF CORRECTING SHAKE RECORDING FOR VIDEO CAMERA}

The present invention relates to a shake correction photographing apparatus for correcting hand shake caused when a digital video camera is photographed. In particular, the present invention relates to a shake correction photographing apparatus that extracts a motion vector of a frame-by-frame image and reduces a size of a moving direction.

The video camera CAMERA continuously records the movement of a subject and records it as an image signal, and shakes such as the movement of the photographer and vibration caused by the surrounding environment, which are generated during shooting, are transmitted to the camera, and the recording is performed by the shake as described above. Movement or trembling occurs in the moving image.

When a motion vector is detected by digitally processing the video signal recorded in the camera, and a video signal compensated for minimizing the detected motion vector is output, the shake generated during the shooting is corrected.

Conventionally, when photographing a moving image (IMAGE), the camera is fixed by using a fixed stand or a tripod, and the camera is photographed by generating a minimum vibration or movement in accordance with the movement of the subject.

The digital video camera is gradually reduced in size and light weight due to the development of technology, and when the camera is held by one hand and is moved while the subject is being photographed, the vibration caused by the photographer's breathing, the vibration caused by the pulse, and the surroundings of the shooting location The vibration caused by the environment is transmitted to the camera and the shake is recorded in the subject image signal photographed.

In other words, even when the subject is not moving, there is a problem that the subject is photographed as the subject is moved by the shaking of the camera.

An object of the present invention is to provide a shake correction photographing apparatus that detects a motion vector of a signal photographed by a digital video camera on a frame basis and corrects the motion vector value of the frame unit to a minimum to reduce a shake occurrence.

In order to achieve the above object, the present invention includes a discrete time conversion unit for converting a digital signal input as a video signal into a video signal of the frequency domain and outputs; A quantization unit for quantizing the video signal applied from the discrete time conversion unit by a control signal; A compression unit which outputs the signal of the quantum unit as a video signal compressed by variable length conversion; An inverse quantum unit for inversely quantizing a signal of the quantum unit; An inverse discrete time conversion unit configured to inversely discrete time convert the frequency signal of the inverse quantum unit and output the digital signal; A motion estimation unit for estimating a motion vector by comparing the inverse discrete time conversion unit signal with an input image signal; A compensator configured to calculate and output a value for compensating for the shaking of the video signal applied from the inverse discrete time conversion unit based on the motion vector value of the motion estimation unit; The control unit monitors and monitors the quantum transformation of the quantum unit, and when the motion vector is compensated for, the control unit for outputting the intra operation mode control signal when the motion vector is not compensated for is characterized by including a configuration.

Hereinafter, the shake correction photographing apparatus according to the present invention will be described with reference to the accompanying drawings.

1 is an explanatory diagram of motion vector detection according to the present invention, FIG. 2 is a functional diagram of a shake correction photographing apparatus using the motion vector of the present invention, and FIG. Illustrative diagram of camera shake correction.

The motion vector is a direction and distance in which a specific pixel (PIXEL) or a block of a certain size (BLOCK) constituting a video signal in a frame unit moves in a previous frame (PREVIOUS FRAME) and a current frame (CURRENT FRAME). Indicates a value.

Referring to FIG. 1, when the A block in the previous frame moves from the current frame to the B block, the motion vector detects the movement distance and the moving direction between the A block and the B block.

In processing the digital video signal, when the motion vector is used as described above, the data amount or size of the transmitted video signal can be reduced and the motion between the frames can be predicted.

Referring to FIG. 2, the shake correction photographing apparatus using the motion vector of the present invention includes a subtractor that calculates and outputs a difference value between an image signal input from a video camera and an image signal applied from a compensator 100 to be described later. 10,

A first switch 20 for outputting an image signal applied from the subtractor 10 by an INTRA or INTER operation mode control signal of a controller 130 to be described later;

A discrete time converting unit 30 converting the digital data video signal applied from the first switch 20 into a video signal in a frequency domain and outputting the converted video signal;

A quantum unit 40 for converting a signal converted into an image signal in the frequency domain from the discrete time conversion unit 30 into a quantized signal by dividing the signal into a predetermined value by a control signal of the controller 130;

A compression unit 50 for outputting a video signal applied from the quantum unit 40 as a compressed video signal by a variable length conversion process;

An inverse quantum unit 60 for inversely quantizing and restoring an image signal applied by being quantized from the quantum unit 40;

An inverse discrete time conversion unit 70 which converts the frequency signal of the inverse quantum unit 60 into inverse discrete time and restores and outputs the digital signal;

An adder 80 which adds and outputs a signal applied from the inverse discrete time conversion unit 70 and a signal applied from the second switch 120 to be described later;

A buffer 90 for temporarily storing and outputting a frame-by-frame video signal applied from the adder 80;

A motion estimator 110 for estimating a motion vector by comparing the video signal applied from the buffer 90 with the video signal input and photographed;

A compensator 100 for compensating for minimization of the shaking or movement of the video signal in units of frames by the video signal applied from the buffer 90 and the motion vector value applied from the motion estimation unit 110;

Supervising and controlling the respective functional units, controlling and monitoring the quantization conversion state of the quantum unit 40, and applying motion vector values by means of an average operation of the motion vector values to compensate for the motion or not. A control unit 130 for analyzing and judging the motion vector according to the analysis judgment, and outputting an INTER operating mode control signal for compensating the motion vector and an intra operating mode control signal for compensating for the motion vector;

The second switch 120 outputs an image signal applied from the compensator 100 to the adder 80 according to an intra or inter operation mode control signal applied from the controller 130.

Hereinafter, a shake correction photographing apparatus using a motion vector according to the present invention having the above configuration will be described in detail with reference to the accompanying drawings.

Since the subtractor 10 for inputting an image signal photographed from the video camera is not initially applied to the compensator 100, the subtractor 10 applies the input signal to the first switch 20 as it is.

The first switch 20 outputs an image signal applied from the subtractor 10 to a discrete time transform transformer (DCT) 30 and at the same time the INTRA of the controller 130. Alternatively, the controller may output the motion estimation unit 110 to the motion estimation unit 110 or not by the INTER operation mode control signal. The controller 130 averages a motion vector value detected by the motion estimator 110 for a plurality of frames inputted for a predetermined period, and the calculated motion vector value of the currently input frame is the averaged calculation process. If the value is larger than the set range, it is controlled in the intra operation mode that does not compensate for the motion vector, in order to reduce the amount of data when it is compressed, and the detected motion vector value is the average. When it is included in the set range compared with the calculated value, the control signal of the INTER operation mode for compensating the motion vector is output.

The discrete time converting unit 30 converts the digital data signal of the video signal into a video signal in a frequency domain and outputs the video signal, so that the video signal is distributed and converted into a video signal in the frequency domain, and the video signal is converted into the frequency domain. Is mainly concentrated or biased in the low frequency band.

The video signal distributed and transformed into the frequency domain as described above is applied to the quantum unit 40 and quantized because it is divided into a corresponding value by the control of the controller 130, and the high frequency band is blocked or blocked by the quantization as described above. The video signal quantized as described above is applied to the compression unit 50 and processed by a variable length conversion (VLC) algorithm (ALGORITHM), and thus is output as a compressed video signal.

In addition, since the quantized signal from which the high frequency portion is removed from the quantum portion 40 is applied to the inverse quantum portion 60 and multiplied by a predetermined value, the high frequency band removed by the processing of the quantum portion 40 is restored. In addition, since the inverse discrete time is applied to the inverse discrete time conversion unit 70, the image signal converted into the frequency domain by the discrete time conversion unit is reconstructed into a digital data signal.

The video signal output from the inverse discrete time conversion unit 70 is a video signal restored to the same signal state as the input video signal photographed and input, and the signal of the compensator 100 is added by the adder 80. And an addition operation is output to the buffer 90 as a frame unit video signal, and there is no motion vector compensation signal output from the compensator 100 at the beginning of the video signal processing.

The buffer 90 outputs the stored image signal in the unit of frame to the compensator 100 and the motion estimation unit 110, and the motion estimation unit 110 is the unit of the frame in the current order applied from the buffer 90. The video signal is compared with the next frame-by-frame video signal applied from the first switch 20, and a motion vector (MV: MOTION VECTOR) is obtained in the pixel (PIXEL) or the assigned block (BLOCK) unit. Estimate and extract the output.

The compensator 100 receiving the motion vector detected from the motion estimation unit 110 and the frame unit video signal from the buffer 90, respectively, compensates or corrects the motion vector to have a small value, thereby subtracting the subtractor 10. ) And the second switch 120.

The second switch 120, the compensation or correction image signal applied from the compensation unit 100 by the INTRA operating mode or INTER operating mode control signal of the control unit 130, A switching process which outputs to the adder 80 or does not output.

The subtractor 10 subtracts the compensating or compensating video signal applied by the compensator 100 from the frame unit video signals input in the following order and outputs the compensating or compensating video signal to the first switch 20. As described above, the motion correction process using the motion vector is repeated to output the corrected and controlled image signal such that the motion vector due to the shake is at the minimum value.

With reference to the accompanying Figure 3, according to the present invention, when explaining the image pickup device shake correction method using a motion vector, before the present invention is not applied, to shoot a point A and a straight line to the C point to shoot In the process, the point B is photographed by the shaking of the camera, which is a photographing apparatus, and when the frame unit video signal photographed at the point B is displayed, the motion vectors are displayed in the order of frames by the large A, B, and C points, There was a problem that the shaking of the video signal is severely displayed.

In the frame unit of the image signal photographed and input by the camera photographing apparatus, an area larger than the size of each frame shown in FIG. 3 is photographed, and an area output by the correction process as described above is outputted and displayed. do.

Accordingly, when the present invention as described above is applied, the motion vector of the B-point frame video signal due to shaking is reduced and corrected as the position of the B 'frame. As shown in the right side of FIG. The movement of the point is corrected and displayed close to the straight line.

According to the present invention having the above-described configuration, the motion vector of the frame-by-frame video signal input by the camera is reduced and output as a corrected video signal, and the shaking due to the movement is minimized. There is an industrial use effect to increase the.

In addition, there is a convenient effect in use to minimize the shake caused by the effect of the movement or vibration to the frame-by-frame image signal taken by the video camera to easily remove.

1 is an explanatory diagram of motion vector detection according to the present invention;

2 is a functional configuration of the camera shake correction device using a motion vector of the present invention,

3 is an explanatory diagram of camera shake correction according to the present invention.

** Explanation of symbols on the main parts of the drawing **

10: subtractor 20: first switch

30: discrete time conversion unit 40: quantum unit

50: compression portion 60: inverse quantum portion

70: inverse discrete time conversion unit 80: addition unit

90: buffer 100: compensation unit

110: motion estimation 120: second switch

130: control unit

Claims (4)

A discrete time conversion unit converting the digital signal input as the video signal into a video signal in the frequency domain and outputting the converted video signal; A quantization unit for quantizing the video signal applied from the discrete time conversion unit by a control signal; A compression unit for outputting a signal of the quantum unit as a video signal compressed by a variable length conversion; An inverse quantum unit for inversely quantizing the signal of the quantum unit; An inverse discrete time converting unit converting the frequency signal of the inverse quantum unit by inverse discrete time conversion and outputting the digital signal; A motion estimation unit for estimating a motion vector by comparing the inverse discrete time conversion unit signal with an input image signal; A compensator for calculating and outputting a value for compensating for the shaking of the image signal applied from the inverse discrete time conversion unit based on the motion vector value of the motion estimation unit; A control unit configured to control and monitor the quantum transformation of the quantum unit and to output an inter operation mode control signal when the motion vector is compensated for and an intra operation mode control signal when the motion vector is not compensated for Calibrated photographing device. According to claim 1, A subtractor for subtracting the compensated motion vector value applied by the compensation unit from the input video signal and outputting the minimized motion vector value; A first switch receiving the video signal of the subtractor and selectively outputting the video signal of the subtractor to the discrete time converting unit and the motion estimation unit according to an intra or inter mode operation control signal of the control unit; A second switch receiving the image signal of the compensator and selectively outputting the image to the adder by an intra or inter mode operation control signal of the controller; An adder configured to add and output an image signal applied from the inverse discrete time conversion unit and an image signal applied from the second switch; And a buffer for temporarily recording and storing the video signal applied from the adder and outputting the buffer to the compensator and the motion estimator, respectively. delete delete