JPH09261530A - Video recorder - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce the circuit scale by using a memory and a motion vector detection circuit in common. SOLUTION: An image signal from a segmentation image selection circuit 21 is fed to an MPEG encoder. The MPEG encoder compresses a received image signal and provides an output of the compressed signal and stores its decoded image to a memory 16 as a reference image. A motion compensation circuit 9 obtains a motion of the reference image to compensate the motion of the reference image. An image frame motion discrimination circuit 22 discriminates whether or not the entire image is in motion based on a motion vector obtained by the motion compensation circuit 9 to obtain an image motion vector. A camera-shake discrimination circuit 24 discriminates the image motion vector to be resulted from a camera-shake, then corrects an image position of a segmented frame of the segmentation image selection circuit 21 based on the image motion vector. Thus, the image signal subjected to camera-shake correction is obtained from the segmentation image selection circuit 21. The memory and the motion detection circuit for correcting camera-shake are used in common for circuits in the MPEG encoder and then the circuit scale is reduced.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a video recording apparatus suitable for performing MPEG standard compression processing.

[0002]

2. Description of the Related Art Conventionally, in a portable video recording device such as a consumer video camera, handheld photographing has become easy due to the downsizing and weight reduction of the device. However, in hand-held photography, the device often vibrates at about 20 Hz or less due to camera shake, and as the magnification of the lens increases, the disturbance of the screen becomes a problem especially in the case of telephoto photography. Therefore, a device having a camera shake correction function has been commercialized.

The camera shake correction function is realized by detecting the camera shake and correcting the camera shake. Regarding the detection and correction of camera shake, there are a method of mechanically detecting and correcting, a method of electrically detecting and correcting, and a method of combining these. However, the method of mechanically detecting and correcting is disadvantageous in terms of downsizing and weight saving, and as a portable type, a method of electrically detecting and correcting is often adopted.
Further, in recent years, digital processing of images has become widespread, and a method of detecting a camera shake by obtaining a motion vector of a video signal by digital signal processing may also be adopted in camera shake correction.

In this case, when the image is digitized, the code amount becomes large in a general picture. Therefore, high-efficiency coding such as MPEG (Movig Picture Experts Group) is applied to the image data. High efficiency coding technology
In order to improve the efficiency of digital transmission and recording,
The image data is encoded at a low bit rate.

FIG. 5 is a block diagram showing a conventional video recording apparatus which has such a camera shake correction function and performs high efficiency coding.

An input image signal from an image pickup device (not shown) is supplied to an image stabilization circuit 2 via an input terminal 1. The size of the effective screen of the image signal from the imaging device is larger than the size of the effective screen of the display device (not shown). The camera shake correction circuit 2 sets a clipping frame having a size corresponding to the valid screen of the display device on the valid screen of the input image signal, and outputs the image signal of the screen of the trimming frame (cutout screen). By correcting the position of the clipping frame according to the camera shake, the image signal with the camera shake corrected is obtained.

The camera shake correction circuit 2 is composed of a memory 3, a motion vector detection circuit 4 and a camera shake correction section 5. The input image signal is delayed by the memory 3 for one frame period, for example. The motion vector detection circuit 4 has an input terminal 1
The image signal of 1 frame before is input from the memory 3 and the image signal of the previous frame from the memory 3, and a motion vector is detected for each block at the four corners of the cut-out screen and output to the camera shake correction unit 5.

The camera shake correction unit 5 corrects the camera shake by cutting out the screen with a cutout frame based on the motion vector. That is, the camera shake correction unit 5 determines whether or not the motion of the blocks at the four corners of the cutout screen is due to camera shake by comparing the motion vectors sequentially input for each frame. Image stabilization unit 5
When it is determined that a motion vector due to camera shake is input, corrects the clipping frame set in the effective screen of the image signal from the memory 3 based on the motion vector,
The image signal of the cut-out screen is output as the output of the image stabilization circuit 2.

In this way, the image compression circuit 6 is provided with the image signal whose image blur is corrected by the image blur correction circuit 2. The image compression circuit 6 is compatible with the MPEG standard. In addition to intra-frame compression for encoding in the same screen, inter-frame compression for compressing the difference between images of two frames by utilizing the correlation between frames. Has been adopted. High compression is possible because the image is also compressed by using the correlation in the time direction.

Further, in the inter-frame compression, if the difference between the previous frame or the subsequent frame and the current frame is simply obtained, the difference becomes large when there is a motion in the image. Therefore,
The motion vector is detected by obtaining the position of the previous or subsequent frame corresponding to the predetermined position of the current frame, and the difference is obtained at the pixel position corresponding to this motion vector to perform motion compensation and reduce the difference value. There is.

The image signal input to the image compression circuit 6 is given to the subtractor 7, the terminal b of the switch 8 and the motion compensation circuit 9. The subtracter 7 performs subtraction with the image data of a motion-compensated image one frame before (hereinafter referred to as reference image), and outputs the difference value to the terminal a of the switch 8 as a prediction error. As a result, it is possible to perform the inter-frame coding in which the difference data is coded by utilizing the redundancy of the image between the frames. When the switch 8 selects the terminal b, the image signal is directly supplied to the DCT circuit 10 for intra-frame compression, and when the switch 8 selects the terminal a, the prediction error is supplied to the DCT circuit 10. Interframe compression is performed. The DCT process is performed in units of horizontal and vertical blocks of 8 × 8 pixels, for example.

The DCT circuit 10 converts the image data represented by the spatial coordinate level into a spatial frequency level, and the DC and AC components of the conversion coefficient are sequentially quantized by a quantizing circuit 11 from a low frequency band to a high frequency band of the horizontal and vertical frequencies. Output to. The DCT transform coefficient is quantized in the quantization circuit 11, and the code amount is reduced.

The quantized output is a variable length coding circuit (hereinafter referred to as V
Variable length coded at 12). Huffman coding, for example, is adopted as the variable-length coding, and data having a higher appearance probability is converted into a shorter code to further reduce the code amount. The encoded output from VLC12 is output terminal 18
Is output via.

The reference image used in the inter-frame compression processing is obtained by locally decoding the quantized output. That is, the quantized output is given to the inverse quantization circuit 13 and inversely quantized, and then given to the inverse DCT circuit 14. Reverse D
The CT circuit 14 performs inverse DCT processing on the inverse quantized output to restore the original spatial coordinate component and outputs the original spatial coordinate component to the adder 15. When the prediction error of the output of the subtractor 7 is supplied to the DCT circuit 10, the inverse D
The output of the CT circuit 14 also becomes a prediction error.

The output of the adder 15 is delayed by one frame by the memory 16, then motion-compensated by the motion compensating circuit 9 and returned to the adder 15 via the switch 17. Adder 15
Adds a prediction error to the motion-compensated reference image data from the motion compensation circuit 9 to reproduce the current frame data (local decoded data) and outputs it as a local decoded signal.

On the other hand, the image signal of the current frame and the memory 16
The image data of the previous frame obtained by delaying for one frame period is given to the motion compensation circuit 9. The motion compensation circuit 9 obtains a motion vector by, for example, matching calculation. The motion compensation circuit 9 extracts the data of the corresponding block from the local decoded signal of the memory 16, corrects it according to the motion vector, and outputs it to the subtracter 7 and also to the adder 15 via the switch 17. In this way, the motion-compensated previous frame data is supplied from the motion compensation circuit 9 to the subtractor 7 as a prediction frame, and the DCT processing is performed on the prediction error from the subtractor 7. The motion vector from the motion compensation circuit 9 is given to the VLC 12, is subjected to variable length coding, and is then multiplexed with the coded output and output.

The luminance signal and the color difference signal are compressed by different processing systems. In this case, the sampling clock of the color difference signal is generally different from the sampling clock of the luminance signal. For example, the sampling clock of the color difference signal is 1/4 of the sampling clock of the luminance signal.
, The ratio of the size of the luminance block and the size of the color difference block, which is the unit of DCT processing, is 1: 4. In this case, a macroblock is composed of 6 DCT blocks of 4 blocks of luminance and 1 block of color difference, and the motion vector is detected in macroblock units.

As described above, in the apparatus of FIG. 5, not only the image compression circuit but also the frame memory and the motion vector detection circuit are necessary for camera shake correction, and there is a problem that the circuit scale is extremely large.

[0019]

As described above, in the above-mentioned conventional video recording apparatus, the frame memory and the motion vector detection circuit are necessary for the image stabilization,
There was a problem that the circuit scale was extremely large.

The present invention has been made in view of the above problems, and an object of the present invention is to provide a video recording apparatus capable of reducing the circuit scale by sharing a memory and a motion vector detection circuit. To do.

[0021]

A video recording apparatus according to the present invention differs from a current screen by a cutout screen selecting means for setting a predetermined cutout frame in an input image and outputting an image signal of a screen in the cutout frame. Storage means for storing an image on the screen as a reference image, and motion compensation means for detecting a motion between the image on the current screen and the reference image stored in the storage means to obtain a motion-compensated reference image. A code for the image signal from the cut-out screen selecting means, which performs intra-picture compression for the picture of the current picture and inter-picture compression for prediction error between the picture of the current picture and the motion-compensated reference picture. Encoding means for outputting the encoded data; camera shake detecting means for detecting the motion of the screen due to camera shake based on the motion obtained by the motion compensating means; and based on the detection result of this camera shake detecting means. Is obtained by including a camera shake correction means for changing the screen position of the clipping frame of the cutout window selection means Te.

In the present invention, the input image is encoded by the encoding means and the encoded data is output. The encoding unit stores an image of a screen different from the current screen as a reference image in the storage unit for the inter-screen compression process. The motion compensator obtains a motion between the reference image and the image of the current screen to compensate the reference image. The camera shake detection means detects the movement of the screen due to the camera shake based on the movement obtained by the movement compensation means. The camera shake correction means controls the cutout screen selection means based on the detection result of the camera shake detection means, and the cutout screen selection means changes the screen position of the cutout frame. In this way, the cut-out screen selection means outputs the image signal in which the camera shake is corrected.

[0023]

Embodiments of the present invention will be described below in detail with reference to the drawings. FIG. 1 is a block diagram showing an embodiment of a video recording apparatus according to the present invention.
1, the same components as those in FIG. 5 are denoted by the same reference numerals.

An input image signal from an image pickup device (not shown) is input to the input terminal 1. The size of the effective screen of the image signal from the imaging device is larger than the size of the effective screen of the display device (not shown). The cutout screen selection circuit 21 sets a cutout frame of a size corresponding to the valid screen of the display device at the screen position specified by the shake determination circuit 24, which will be described later, within the valid screen of the input image signal. Output an image signal.

In the present embodiment, the clipping frame is set twice for the input image signal of a predetermined frame. That is, the cutout screen selection circuit 21 first sets the cutout frame at the same position as the cutout frame for the input image signal of the previous frame in order to detect the shake, and then corrects the shake. The cutout frame is set based on the output.

The image signal of the cut-out screen from the cut-out screen selection circuit 21 is given to the subtractor 7 and the motion compensation circuit 9 which form the MPEG encoder, and also to the screen frame position information detection circuit 23. The structure of the MPEG encoder is the same as that shown in FIG.

That is, also in this embodiment, not only the intraframe compression processing but also the interframe compression processing can be performed on the input image signal. The subtractor 7 receives the motion-compensated reference image data from the motion compensation circuit 9 described later, obtains the difference between the image data of the current frame from the cut-out screen selection circuit 21 and the reference image data, and predicts the difference value as a prediction error. Is output to the terminal a of the switch 8. By using this prediction error, it is possible to perform inter-frame coding in which differential data is coded using the redundancy of images between frames.

The switch 8 is controlled by a picture signal from the encoder mode switching control circuit 26. That is, the switch 8 selects the terminal b when the intra-frame compression processing is designated by the picture signal and supplies the image data of the current frame to the DCT circuit 10 as it is, and when the inter-frame compression processing is designated. Selects the terminal a and supplies the prediction error to the DCT circuit 10.

The encoder / mode switching control circuit 26 is supplied with the output of the camera shake correction instruction circuit 25. The camera shake correction instruction circuit 25 sets the camera shake correction mode in which the camera shake correction is performed based on the user operation to the encoder / mode switching control circuit 26.
Output to instruct. The encoder / mode switching control circuit 26 is used for the I picture by intra-frame compression and the P for inter-frame compression in the MPEG encoder.
Various modes that define the method of creating a picture or B picture are determined, and picture signals for creating I, P, and B pictures are output in correspondence with these modes. In the present embodiment, the encoder / mode switching control circuit 26 outputs a picture signal corresponding to the mode in which the cycle of the I picture or P picture is M = 1 when the camera shake correction mode is designated. ing. In the mode of M = 1, a B picture obtained by interframe compression using a temporally backward image as a reference image is not created. This ensures the detection of camera shake.

The DCT circuit 10 receives the image data or its prediction error from the switch 8, converts the image data represented by the spatial coordinate level into the spatial frequency level, and converts the DC and AC components of the conversion coefficient into horizontal and vertical frequencies. The signals are sequentially output to the quantization circuit 11 from the low band to the high band. The quantizing circuit 11 quantizes the input DCT transform coefficient using a predetermined quantizing coefficient. The larger the horizontal and vertical high frequencies of the DCT transform coefficient, the larger the quantization coefficient is set. The quantized output from the quantization circuit 11 is given to the VLC 12. The VLC 12 employs, for example, a Huffman table created based on the appearance probability of the data of the pair of non-zero data and zero run of the quantized output, and performs the Huffman coding. As a result, a shorter code is assigned to data having a higher appearance probability, and the data amount is further reduced. The encoded output from VLC 12 is output via output terminal 18.

On the other hand, the quantized output from the quantization circuit 11 is
It is also provided to the inverse quantization circuit 13 to obtain the reference image data. The inverse quantization circuit 13 multiplies the input quantized output by a quantization coefficient to restore the data before quantization, and outputs the data to the inverse DCT circuit 14. The inverse DCT circuit 14 outputs DC to the inverse quantized output.
The inverse processing of the T circuit 10 is performed to restore the spatial frequency component to the original spatial coordinate component and output it to the adder 15.

The output of the adder 15 is delayed by the memory 16 for one frame period, and then the motion compensation circuit 9 and the switch 17 are provided.
Is given to the adder 15 via. That is, the prediction error is supplied from the inverse DCT circuit 14 to the adder 15, and the reference image data of one frame before is given through the switch 17, and the adder 15 adds the two inputs to add the current frame. Restore the data. The locally decoded data from the adder 15 is given to the memory 16. memory
16 holds the local decoded data from the adder 15 and outputs it to the motion compensation circuit 9.

On the other hand, the image data of the current frame from the cutout screen selection circuit 21 is also given to the motion compensation circuit 9. The motion compensation circuit 9 is supplied with image data of one frame before and after, and detects the motion of the image of the current frame in macroblock units. Further, the motion compensation circuit 9 determines the blocking position of the reference image stored in the memory 16 based on the motion vector and outputs the motion-compensated reference image block data. In the present embodiment, the motion compensation circuit 9 is
While outputting the motion vector detected using the output of the cutout screen selection circuit 21 at the time of camera shake detection to the screen frame motion determination circuit 22, by the motion vector detected using the output of the cutout screen selection circuit 21 at the time of camera shake correction, The reference image block data is motion-compensated and output to the subtractor 7 and the switch 17.

The screen frame position information detection circuit 23 detects the screen position of the cutout frame selected by the cutout screen selection circuit 21 and outputs it to the screen frame movement determination circuit 22 as screen frame position information. The motion compensation circuit 9 also supplies the screen frame motion determination circuit 22 with motion vectors at the time of detecting a shake of a plurality of macro blocks. The screen frame motion determination circuit 22 compares the motion vectors of a plurality of macroblocks output from the motion compensation circuit 9 and determines that the preset number or more of motion vectors have substantially the same size and direction. Determines that the entire screen has moved.

For example, in the case where the motion vectors of 10 macroblocks are supplied from the motion compensating circuit 9, there are errors in the magnitude and direction of the motion vectors of 8 macroblocks, that is, variations. If it is within 5%, the screen frame movement determination circuit 22 determines that the entire screen has moved. In this case, the screen frame motion determination circuit 22 outputs a predetermined motion vector of the plurality of motion vectors having substantially the same size and direction to the camera shake determination circuit 24 as a screen motion vector indicating the motion of the screen. It is supposed to do.

The camera shake judgment circuit 24 is a screen frame movement judgment circuit.
By comparing the screen motion vectors sequentially input from 22, it is determined whether or not the screen motion indicated by the screen motion vector is due to camera shake. If the camera shake determination circuit 24 determines that the change in the screen motion vector is due to camera shake, the camera shake determination circuit 24
The screen motion vector from 22 is cut out and output to the screen selection circuit 21. Cutout screen selection circuit 21
Corrects the screen position of the clipping frame based on the screen motion vector.

Next, the operation of the embodiment thus configured will be described with reference to the explanatory views of FIGS. 2 to 4. FIG. 2 shows a cutout frame, a cutout screen, and a local decode screen. 2 (a) to 2 (c) show input images of the nth to (n + 2) th frames, respectively.
2D, 2F, and 2H show cut-out screens for the input images of the nth to n + 2th frames, respectively, and FIG.
(G) shows the local decode screens stored in the memory 16, respectively. FIG. 3 is for explaining the operation of the cut-out screen selection circuit 21.

An input image signal from an image pickup device (not shown) is input to the input terminal 1. This input image signal is supplied to the cut-out screen selection circuit 21. The effective screen of the input image signal is larger than the size of the effective display screen of the display device (not shown).

Now, it is assumed that camera shake correction is not performed.
In this case, the cutout screen selection circuit 21 outputs the image signal of the cutout screen at the fixed predetermined screen position without changing the screen position of the cutout frame.

On the other hand, the encoder / mode switching control circuit 26
Determines a compression mode corresponding to the normal MPEG standard and outputs a picture signal. The picture signal is supplied to the switches 8 and 17. When the I signal is instructed by the picture signal, the switch 8 selects the terminal b. As a result, the block data from the cut-out screen selection circuit 21 is given to the DCT circuit 10 as it is. D
The CT circuit 10 performs DCT processing on the image data, and the quantization circuit 11
Quantizes the transform coefficients to reduce the bit rate. The quantized output is given to the variable length coding circuit 12, is variable length coded, and then output via the output terminal 18.

On the other hand, the quantized output from the quantization circuit 11 is
In order to obtain the reference image, it is given to the inverse quantization circuit 13 and inversely quantized. The inverse quantized output is given to the inverse DCT circuit 14, subjected to inverse DCT processing to be returned to the original image data, and then given to the adder 15. The locally decoded image from the adder 15 is stored in the memory 16 as a reference image. Memory 16
The reference image from is subjected to motion compensation in the motion compensation circuit 9 and supplied to the subtractor 7.

When the P and B pictures are instructed by the picture signal, the switch 7 selects the terminal a and gives the prediction error from the subtractor 7 to the DCT circuit 10.
The subtractor 7 outputs the difference between the output of the cut-out screen selection circuit 21 and the output of the motion compensation circuit 9 as a prediction error. Thus, in this case, the DCT processing for the prediction error,
The quantization process and the variable length coding process are performed, and the coded output is output from the output terminal 18.

Even in this case, the quantization circuit 11
The output of is returned to the original pixel data by the inverse quantization circuit 13 and the inverse DCT circuit 14 and then supplied to the adder 15.
When the P and B pictures are created, the switch 17 is on, and the motion-compensated reference image from the motion compensation circuit 9 is supplied to the adder 15 via the switch 17. Inverse DCT
The prediction error from the circuit 14 and the reference image are added by the adder 15 to restore the image data of the current frame, which is stored in the memory 16 as the image signal of the local decoded screen. After that, the image is encoded by the same process.

Here, it is assumed that the image stabilization instruction circuit 25 gives an instruction for image stabilization. In this case, the encoder / mode switching control circuit 26 specifies the compression mode of M = 1 to prohibit the B picture from being created.

Now, it is assumed that the image signal of the input image 31 of the nth frame shown in FIG. 2A is input. Immediately after the start of the camera shake correction, the cutout screen selection circuit 21 sets the cutout frame 32 at a predetermined screen position set in advance. The cutout screen selection circuit 21 outputs the image signal of the cutout screen 33 (FIG. 2D) surrounded by the cutout frame 32. The compression process for the image signal from the cut-out screen selection circuit 21 is similar to the case where the image stabilization is not performed.
Thus, the image signal of the local decode screen 33 'corresponding to the cut-out screen 33 is stored in the memory 16.

Next, the image signal of the input image 34 (FIG. 2B) of the (n + 1) th frame is input through the input terminal 1. Note that this input image is assumed to be captured in a state in which camera shake in the lower left direction has occurred during image capturing. In this case, the cut-out screen selection circuit 21 first sets the cut-out frame 35 (broken line) at the same position as the screen position of the cut-out frame 32 in the previous frame in order to detect camera shake. The cutout screen selection circuit 21 is a cutout screen corresponding to the cutout frame 35.
The 36 image signals are output to the motion compensation circuit 9. The motion compensation circuit 9 detects the motion of a plurality of macroblocks in the image of the cut-out screen 36 and the image of the local decode screen 33 'stored in the memory 16. The plurality of motion vectors detected by the motion compensation circuit 9 are supplied to the screen frame motion determination circuit 22.

The screen frame motion determination circuit 22 includes a motion compensation circuit 9
If the magnitude and direction of a predetermined number or more of the plurality of motion vectors from 1 to 3 are within a predetermined error range, it is determined that the entire screen has moved. Further, the screen frame position information detection circuit 23 detects the information on the screen position of the cutout frame 35 set by the cutout screen selection circuit 21 and supplies it to the screen frame movement determination circuit 22. In this case, since the entire screen is moving in the upper right direction due to camera shake in the lower left direction, the size and direction of the predetermined number or more of motion vectors are within the error range, and the screen frame motion determination circuit 22 determines The motion vector obtained for the macro block is output to the camera shake determination circuit 24 as a screen motion vector indicating correction of the screen position of the clipping frame 35.

The shake determination circuit 24 is a screen frame movement determination circuit 22.
By comparing the screen motion vectors sequentially input from, it is determined whether or not the screen motion vector is due to camera shake. FIG. 4 is for explaining this judgment. For example, as shown in FIG. 4, when the directions of sequentially input screen motion vectors are not the same direction, that is, when the direction of movement of the entire screen changes in the same manner as camera shake, it is determined that camera shake has occurred. to decide. It should be noted that the camera shake determination circuit 24 may determine whether or not the camera shake is in consideration of the frequency of camera shake. In the example of FIG. 2, the camera shake determination circuit 24 determines that the screen motion vector is due to camera shake,
This screen motion vector is output to the cutout screen selection circuit 21.

The cut-out screen selection circuit 21 changes the screen position of the cut-out frame on the basis of the screen motion vector supplied from the camera shake determination circuit 24. FIG. 3 shows the operation in this case. The cut-out screen selection circuit 21 receives the cut-out frame 35 set at the time of detecting the camera shake and inputs the screen motion vector
It corrects by 40 and sets the cutout frame 37 shown in FIG.2 (b). The cut-out screen selection circuit 21 outputs the image signal of the cut-out screen 38 (FIG. 2 (f)) of the cut-out frame 37 to the MPEG encoder as an output after camera shake correction.

The compression process in this case is similar to that in the case where the camera shake is not corrected, and the image signal of the local decode screen 38 ′ corresponding to the cut-out screen 38 is stored in the memory 16. Further, the motion compensation circuit 9 detects a motion vector between the image of the cutout screen 38 and the image of the local decode screen 33 '.

If the camera shake determination means 24 determines that the camera shake has not occurred, the cutout screen selection circuit 21 sets the cutout frame 35 in the previous frame again, and the cutout screen image signal of the cutout screen 36 is set. Is output.

Thereafter, the same processing is performed to detect and correct the camera shake. For example, it is assumed that the image signal of the image 42 of the (n + 2) th frame shown in FIG. 2C is input. In this case, the cutout screen selection circuit 21 first sets the cutout frame 37 (broken line) and outputs the image signal of the cutout screen 43 in the frame 37 to the motion compensation circuit 9. The motion compensation circuit 9 obtains motion vectors of a plurality of macroblocks between the cutout screen 43 and the local decode screen 38 '.

It is assumed that a shake in the upper right direction occurs in the image 34 of the (n + 1) th frame. In this case,
Since the entire screen of the image 42 of the (n + 2) th frame moves in the lower left direction, the screen frame motion determination circuit 22 causes the motion compensation circuit 9 to move.
A screen motion vector indicating that the entire screen has moved is output to the camera shake determination circuit 24 based on the motion vector from the. The camera shake determination circuit 24 determines that the screen motion vector is due to camera shake, and outputs the screen motion vector for the cutout frame 37 to the cutout screen selection circuit 21.
In this way, the cutout screen selection circuit 21 corrects the screen position of the cutout frame 37 by the screen motion vector and sets the cutout frame 44. As a result, the image signal of the cutout screen 45 (FIG. 2H) of the cutout frame 44 is supplied to the MPEG encoder as the image signal after the shake correction.

As described above, in the present embodiment, M
Image stabilization is performed using the circuit provided in the PEG encoder. Since the frame memory and the motion detection circuit for camera shake correction are shared, the circuit scale can be reduced.

The present invention is not limited to the above embodiment. For example, various methods can be adopted for the camera shake correction method, and instead of the cut-out screen selection circuit, a circuit in which the driving timing of the CCD in the image pickup device is changed so that the cut-out frame can be moved is adopted. It is also possible. Further, the compression mode during camera shake correction is not limited to M = 1.

[0056]

As described above, according to the present invention, the circuit scale can be reduced by sharing the memory and the motion vector detecting circuit.

[Brief description of drawings]

FIG. 1 is a block diagram showing an embodiment of a video recording device according to the present invention.

FIG. 2 is an explanatory diagram for explaining operation of the embodiment;

FIG. 3 is an explanatory diagram for explaining operation of the embodiment;

FIG. 4 is an explanatory diagram for explaining operation of the embodiment;

FIG. 5 is a block diagram showing a conventional video recording device.

[Explanation of symbols]

9 ... Motion compensation circuit, 16 ... Memory, 21 ... Cutout screen selection circuit, 22 ... Screen frame motion determination circuit, 23 ... Screen frame position information detection circuit, 24 ... Hand shake determination circuit

Claims (4)

[Claims] 1. A cut-out screen selecting means for setting a predetermined cut-out frame on an input image and outputting an image signal of a screen in the cut-out frame; a storage means for storing an image of a screen different from the current screen as a reference image; A motion compensating unit for detecting a motion between the image of the current screen and the reference image stored in the storage unit to obtain a motion-compensated reference image; Coding means for performing inter-picture compression with respect to a prediction error between the picture of the current picture and the motion-compensated reference picture, and outputting coded data for the picture signal from the cut-out picture selection means; Camera shake detection means for detecting the movement of the screen due to camera shake based on the movement obtained by the means, and of the cutout frame of the cutout screen selection means based on the detection result of this camera shake detection means. An image recording apparatus comprising: a camera shake correction unit that changes a screen position. 2. The motion compensating means detects a motion vector indicating a motion in a predetermined block unit, and the camera shake detecting means detects a screen position of a cutout frame set by the cutout screen selecting means. Screen frame position detecting means, and a screen frame motion determining means for obtaining a screen motion vector indicating the motion of the entire screen based on the detection result of the motion vector of the plurality of blocks obtained by the motion compensating means and the screen position of the clipping frame When the screen motion vector is determined to be due to camera shake, the camera shake correction unit provides the screen motion vector to the cutout screen selection unit for changing the screen position of the cutout frame. Claim 1 characterized by the above-mentioned.
The video recording device according to. 3. The cut-out screen selection means sets a cut-out frame at the same screen position as the cut-out frame set in the previous screen for detecting a shake for the input image of the current screen, and corrects the shake. The video recording apparatus according to claim 1, wherein a clipping frame is set based on the camera shake correction unit. 4. The video recording apparatus according to claim 1, wherein the encoding unit does not perform the inter-screen compression process using the backward reference image in the image stabilization mode.