WO1999067947A1

WO1999067947A1 - Image sensing/storing method and image sensing device

Info

Publication number: WO1999067947A1
Application number: PCT/JP1998/002804
Authority: WO
Inventors: Hiromi Watanabe; Tetsuya Nakagawa; Yuji Hatano; Yutaka Okada
Original assignee: Hitachi, Ltd.
Priority date: 1998-06-24
Filing date: 1998-06-24
Publication date: 1999-12-29
Also published as: AU7932798A

Abstract

An image sensing device which records an image as digital signals, produces a good picture shot at a perfect moment, and further produces continuous pictures representing a motion of the object during a short time. The intraframe encoding of the signals of a reference image frame generated on depressing a shutter button is performed and the interframe encoding of the signals of image frames before and after the reference image frame in terms of time is performed by utilizing the correlation between the image frames and the reference image frame or the correlation between the image frames and an image frame predicted from the reference image frame. The encoded signals are recorded in a storing means as instantaneous dynamic images. This constitution enables (1) the movement and the change of the facial expression to be recorded, (2) the parts desired to be recorded to be reliably imaged, and (3) the compression efficiency to be improved and a high quality image to be produced.

Description

Description Video imaging storage method and video imaging device

The present invention relates to a video information storage method and a video imaging apparatus, and more particularly, to a method of converting a video signal into a digital signal, compressing and encoding the video signal and storing the video information, and capturing a still image and a short-time video to generate a digital signal. The present invention relates to a video imaging device that records as a video. Background art

2. Description of the Related Art Conventionally, as a device for recording a video as a digital signal, a digital camera for obtaining a still image and a video camera for storing a moving image are well known. When images are stored as digital signals in these devices, the JPEG (Joint Photographic Image coding Experts Group) method, MPEG ( Moving Picture coding Experts Group) 'method is used.

A digital camera that obtains a still image converts an optical image signal input from a camera lens into an analog electric signal, converts the analog electric signal into a digital signal, and cooperates with a camera shutter to generate an image file. The digital signal of the image frame to be recorded is extracted from the frame train, and the digital signal of the extracted frame image is extracted. For example, it is encoded by intra-frame encoding such as the JPEG compression method and stored in a storage device. A conventional digital camera for still images can record a single image frame from this series of image frames at the same time, but the shutter that the photographer desires The video of the dance and the image frame actually selected by the shutter do not always match. In other words, it does not compensate for a one-hour chance.

As a still image camera for compensating for the above-mentioned shutter chance, a digital camera for a still image as shown in FIG. 16 has been proposed in a patent publication (Japanese Patent Laid-Open No. 3-213609).

The video signal captured by the imaging unit Z1 is temporarily stored in the buffer memory Z2. When the shutter is pressed, the image frame at that time is recorded in memory pack Z3, and at the same time, writing to buffer memory Z2 is prohibited. Next, the image frame before the shutter chance stored in the buffer memory Z2 is stored in the memory pack Z3 via the bus B. Further, depending on the shooting mode, the image frames after the shutter chance are also stored in the memory pack Z2 via the bus A.

However, since such digital cameras aim to record a plurality of image frames in order to capture the image frame at the moment of a single evening, each image frame is stored as it is. Force recorded on the device It is only a device that encodes each image frame in a frame such as JPEG. As a result, the compression ratio is not increased because the correlation between the image frames is not used. However, if the number of image frames to be captured and stored is increased to compensate for the shutter chance, the capacity of the storage device must be increased.

In some cases, an imaging storage device wants to record instantaneous changes in human facial expressions and movements.However, conventional digital cameras are used to obtain still images, so instantaneous changes within a few seconds are independent. It is not possible to capture, store, and play back images. Even if a conventional digital camera performs continuous shooting on a subject and expresses the movement, the conventional still camera captures one image frame, compresses and encodes the image signal, and stores it. Since the sequence stored in the device is taken, the next image frame cannot be shot until the image frame shot in the storage device is recorded. A video camera capable of capturing and recording moving images is a device that compresses and records a series of videos using interframe predictive coding such as MPEG, and can express motion by recording continuous image frames. However, there is a problem that an editing operation is required in order to take out and use a desired image frame from the recorded continuous image frames. In addition, although the MPEG system has excellent performance as a moving image compression method, it is considered that instantaneous moving images may have different degrees of importance for multiple image frames included in the moving images. It has not been.

Disclosure of the invention

Therefore, a first object of the present invention is to accurately capture a video of a one-shot video, record and reproduce the video, and obtain a necessary image frame memory. An object of the present invention is to provide a video image pickup device such as a digital camera which can be configured with a small capacity.

Another object of the present invention is to provide a video image capturing and storing method and a video image capturing apparatus for recording the movement of a subject and changes in facial expressions as independent instantaneous moving images.

In order to achieve the above object, a video imaging / accumulating method according to the present invention converts an optical video signal into a series of image frames of a digital signal, and converts the series of image frames into a series of one or more shutters. , One of the extracted image frames is intra-coded as a basic frame, and a reference image frame (for example, a shutter) is stored in a plurality of stored image frames. (The image frame of one evening), and the image frames before and after the reference image frame are correlated with the reference image frame, or the image frames predicted from the reference image frame. Inter-frame encoding is performed using the correlation between and, and the encoded signal is recorded in the storage device.

Further, in the image pickup and accumulation of the present invention, a photoelectric conversion unit that changes an optical image signal into a series of image frame trains, a shutter, and a shutter are pressed from the series of image frame trains (hereinafter referred to as “shutter”). Means for extracting a plurality of image frames before and after, and intra-frame encoding using one of the extracted image frames as a basic frame, and using the plurality of image frames as a reference The image frame temporally before and after the image frame is correlated with the reference image frame, or is predicted from the reference image frame. It comprises an encoding unit for performing inter-frame encoding using the correlation with the measured image frame, and a storage device for storing the encoded signal.

In a preferred mode, the basic frame is configured so that the image frame of the “Shutdown-Chance” is selected. BRIEF DESCRIPTION OF THE FIGURES

FIG. 1 is a block diagram showing an embodiment of an image capturing apparatus according to the present invention.

FIG. 2 is a block diagram showing a detailed configuration of a main part of FIG. FIG. 3 is an explanatory diagram of a recorded image and a compression method of an embodiment of the image capturing apparatus according to the present invention.

FIG. 4 is a signal processing time chart of an embodiment of the image capturing apparatus according to the present invention.

FIG. 5 is an explanatory view showing an image frame and a block diagram of an embodiment of a shadow image pickup device according to the present invention.

FIG. 6 is a processing flowchart (sequence) of an embodiment of the image capturing apparatus according to the present invention.

FIG. 7 is a processing flowchart (a part of a picture) of an embodiment of the image capturing apparatus according to the present invention.

FIG. 8 is a processing flowchart of an embodiment of the image capturing apparatus according to the present invention.

FIG. 9 is a diagram showing an accumulation structure of image data of a digital camera in image imaging according to the present invention. FIG. 10 is a block diagram showing a second embodiment of the image capturing apparatus according to the present invention.

FIG. 11 is a block diagram showing a main configuration of FIG. 10. FIG. 12 is an explanatory diagram of a recorded video and a compression method of the second embodiment of the image capturing apparatus according to the present invention.

FIG. 13 is a processing time chart of the second embodiment of the image capturing apparatus according to the present invention.

FIG. 14 is an explanatory diagram of a recorded video and a compression method of the second embodiment of the image capturing apparatus according to the present invention.

FIG. 15 is a block diagram showing a configuration of a digital camera according to the present invention.

Fig. 16 is a block diagram showing the configuration of a conventional digital camera that compensates for photo opportunities. BEST MODE FOR CARRYING OUT THE INVENTION

Embodiment 1>

FIG. 1 is a diagram showing a configuration of an embodiment of a shadow image capturing apparatus according to the present invention. As shown in the figure, a lens 1 that captures an image as an optical signal, an image sensor 2 that converts the optical signal into an electric signal, a signal converter 3 that converts the electric signal into, for example, a luminance signal and a chrominance signal, and its luminance An AZD converter 4 for converting color signals into digital signals, a release and shutter button 5 for capturing images and inputting a shutter point as a signal 5, and a mode (e.g., a video capture cycle, Compression coding mode, at the time of the reference image frame Mode input unit 6 for setting the image data, a general control unit 7 for creating control signals for the entire device, an image signal capture switch 8, an image frame memory 9 for temporarily storing captured images, Shutter switch 10 for selectively reading image frames from frame memory 9, Compression encoder 11 for performing compression encoding of selectively read image frame signals, Compression-encoded signals And a storage control device 14 for controlling the storage of the storage device 14. It also has an image playback unit. The image playback unit monitors the video to be captured and displays the stored video, a signal converter 20 that converts it into a monitor display signal, and a digital video signal. It comprises a DZA converter 19 for converting to an analog signal, a selector 18 for selecting a decoded video or an input video, a frame memory for decoding 17, a decompressor 16, and a decoding code buffer 17.

FIG. 2 shows the configuration of a characteristic part of the first embodiment, which is the encoding and decoding part of the signal surrounded by the dotted line in FIG. Image frame memory 9 is used to store image frames in any part of the frame memory 94-999 (FM 1 power, FM6, etc.) for 6 image frames, and the frame memory 94 ... 99. It consists of a selector 92 for writing or selecting, and a selector 93 for selecting a frame to be read from the frame memory 94-99, and a memory controller 91 for controlling the selectors 92 and 93. You. Further, the compression encoder 11 is a block encoder B 1 that decomposes the image frame captured via the switch 10 into blocks, and performs intra-frame encoding on the signal of the block. The output of the selector B 2 for selecting whether to perform inter-frame coding, the discrete cosine converter B 3 for performing a discrete cosine transform of the signal of the selected block, and the output of the discrete cosine converter B 3 are quantized. Quantizer B 4, a variable-length encoder B 5 that performs variable-length encoding of the quantized signal, an inverse quantizer B 6 that inversely quantizes the quantized signal, and an inverse quantizer B 6 , An inverse discrete cosine converter B7 that performs an inverse discrete cosine transform of the output signal, and a selector B8 and a selector B8 that select encoded signals corresponding to intra-frame and inter-frame encoded signals. An adder BC for adding the selected signal and the output from the inverse discrete cosine converter B7 to create a decoded frame, a memory B9 for storing the signal of the decoded frame, a memory B9, and a blocking unit B A motion detector BA that detects block motion using the signal from It consists of subtractor BB for calculating a difference between frames of the motion compensated referenced block signals from the signal and the motion detector from the lock equalizer B 1.

FIG. 3 is a view for explaining a compression method and a recorded image frame in the digital camera device of FIG. 1 which is the first embodiment.

In this embodiment, in order to compress an image signal, a reference image frame is intra-frame coded and recorded at the head of a compressed bit stream for recording, and an image frame temporally before and after the reference image frame is recorded. The frame performs inter-frame coding (prediction coding) from the reference image frame or inter-frame coding from an image frame predicted from the reference image frame, and performs bit coding of the reference image frame. It is recorded after the toast stream. In particular, the reference image frame is an image frame selected by a shutter chance of the digital camera device.

Figure 3 (a) shows a series of image frames i l, f 2

f 20 is simulated in time. Image frame ί10 is the image frame selected by the shutter, and the image frames f 6, f 7 ′, f 9, and the time before the image frame f 10 The subsequent image frames f 11, f 12,..., F 14 are also compression-encoded and stored in the storage device 14 in the digital camera.

FIG. 3 (b) is a diagram for explaining a form of compression encoding of a recorded image frame; f6, f7,... F14.

In the compression encoding, first, the image frame f 10 is used as a reference image frame, and intra-frame encoding is performed.The image frame f 6 that is temporally earlier than the reference image frame f 10 is used as the reference image frame f Predictive coding (interframe coding) is performed starting from 10. The image frames f 7, f 8 and f 9 are predictively coded (inter-frame coding) from the image frames f 10 or f 6. The image frame temporally later than the reference image frame f 10 ί 1 1 ··· f 14 is the first image frame; f 14 is the predictive coding from the reference image frame f 10 (Interframe coding), and the image frames f11, fl2, and fl3 are predictively coded (interframe coding) from the image frame flO or fl4.

FIG. 4 shows the digital camera device of FIG. 1 described in FIG. FIG. 4 is a processing time chart when performing compression encoding. In the figure, the number enclosed by a rectangle indicates the number of the image frame written to the memory 9, and the number enclosed by the circle indicates the number of the image frame read from the memory 9.

The input image frames shot at the video capture period set in the mode input section 6 are stored in the frame memory while the release signal from the release and shutter button 5 is active (low level in the figure). Import to 9. In the figure, the third image frame is stored in memory 94 (FM1), the fourth image is stored in memory 95 (FM2), the fifth image is stored in memory 96 (FM3), and the sixth image is stored in memory 97 (FM7). FM 4), number 7 to memory 98 (FM 5.), number 8 to memory 9.9 (F 6), and number 9 again to memory 94 (FM 1). . When the shutter button is pressed during the 10th image frame, the 10th image frame is loaded into memory 95 (FM2), and the 11th image frame is loaded into memory 96 (FM3). At the same time, the 10th image frame is read out from the memory 96 (FM2) and is intra-coded as a reference image frame. At the same time, the image frame is used as a reference image frame and is used as a frame memory ( Write to B 9) in FIG. 2 and simultaneously write the compressed code to the code buffer 12. The written code is recorded in the storage device E by the storage control device D at an appropriate timing. The image frame No. 12 is taken into the memory FM 2 where the reference image frame is stored. At the same time, note Re-read the sixth image frame captured in FM4 and perform predictive encoding (inter-frame encoding) from the reference image frame.

At the same time, the compression code is written to a code buffer 12, and the image frame is written to a frame memory (B9 in FIG. 2) as a reference image frame. 13 The image frame No. 3 is taken into the memory FM4 from which reading has been completed. At the same time, the 7th image frame is read from the memory FM5, and the predictive coding from the 10th or 6th image frame, which is the reference image frame in the frame memory (interframe code) ). The result is output to the sign buffer 12. Similarly, process the 8th image frame while capturing the 14th image frame. Subsequently, the processing of the ninth image frame is performed, and the compression encoding processing of the image frame when the shutter button is pressed and before that is completed.

Next, the No. 14 image frame captured in the memory FM No. 5 is subjected to predictive coding (inter-frame coding) from No. 10 which is the reference image frame, and the code is stored in the code buffer. At the same time as outputting to C, it is replaced with the sixth image frame and written into the frame memory. Next, the 11th image frame is read out from the memory FM3, the prediction encoding (interframe encoding) is performed from the 10th or 14th image frame, and the code is output to the code buffer 12. . In the same way, the processing is performed with steps 12 and 13, and the processing of the image frames after the shutter button is pressed is terminated.

The processing of the compression encoding unit according to the first embodiment is described with reference to FIGS. This will be described in more detail with reference to FIG.

FIG. 5 is a diagram showing image frames to be compression-encoded and blocks of processing units. The upper part of the figure is the same as Fig. 3 (b), which is the same as a series of image frames ί6, f7 · · · · · f14 recorded in the memory 9, and for convenience of the following description, Each P ic (

— 4), P i c (— 3) P i c (4). That is, the image frame f10 of the shutter chance is represented as Pic (0), and the image frame earlier in time is defined as minus (one) (+), and the temporally earlier image frame is represented as Pic (0). Subsequent image frames are represented by a plus. Each image frame is decomposed into macroblocks called macroblocks (MB) of 16 x 16 pixels. One image frame is composed of horizontal M macroblocks and vertical N macroblocks (M and N are integers). The macro block is further divided into two sections each horizontally and vertically, and is composed of four blocks of 8 pixels x 8 pixels.

FIG. 6 is an operation flowchart of compression encoding according to the first embodiment. When the release button is pressed, the processing is started (600), and in the "waiting" state (601), the image frames are sequentially taken into the frame memory 9. When the shutter button 5 is pressed (602), a sequence header containing the entire information of the video to be recorded is first written (603).

Next, the number of the image frame to be encoded is determined (604). The first is a picture frame of a one-hour shot, Pic (10), followed by Pic (—4), Pic (—3), and Pic (—2). , Pic (—1) and the picture frame force before the shutter chance is determined, and then Pic (4), Pic (1), Pic (2), Pic (3) and shutter The number after the license is determined. Picture processing with the determined number as the argument M um

(PictP (Mum)) is performed.

FIG. 7 shows a processing flowchart of the picture processing (PictP (Mum)).

Picture processing is processing in units of image frames. First, information specific to the image frame is written as a picture header (701). Next, an image frame memory in which the image frame of the number Mum passed as the argument is stored is selected, the image data is read from the image frame (702), and the image data is decomposed into the above-described macroblock (7). 0 3), Read one macro block at a time. The argument M um is examined. If M um = 0, the frame is a reference image frame, so the coding is performed by intra-frame coding (704). This is performed by intraframe encoding. In the figure, macroblock intra-frame encoding is performed by setting the TYPE signal indicating the type of macroblock to INTRA. If the argument Mum is -4 or 4, predictive coding from the reference image frame Pic (0) is performed. In predictive coding, motion detection is performed (706, 707), and the macroblock most similar to the macroblock to be coded is cut out from the reference image frame as a reference macroblock. Macroblock to be encoded When the sum of the absolute values of the difference values is small to some extent, the difference value is used to perform compression processing (interframe coding). If the difference value becomes so large that the use of the difference does not increase the compression effect, the macroblock to be coded is directly coded (intra-frame coding). In the figure, the processing is performed in the TYPE determination part (708). If the absolute value sum of the difference values is small, TYPE is set to INTER. Otherwise, TYPE is set to INTRA. If the argument Num is —3, 1 2 and 1 1, the reference image frame is composed of two image frames, the reference image frame Pic (0) and the image frame 14 encoded from the prediction. Predictive encoding from frames is performed.

First, a macroblock (RefMB0) most similar to the reference image frame Pic (0) is detected. Next, the most similar macroblock (RefMBL) is detected from the image frame Pic (14). The difference between each reference macroblock (Ref MB0) (Ref MB1) and the difference between the macroblock that is the average of Ref MBO and Ref MB1 is calculated, and the absolute value of the difference value is calculated. When the sum of the values is small to some extent as described above, compression processing (inter-frame encoding) is performed using the difference value. Otherwise, the encoding target macro block is encoded as it is. In the figure, the processing is performed in the TYPE determination part 708.

The TYPE determination unit 708 performs another important determination. If it is determined that the encoding target macroblock and the reference macroblock are the same or that it can be determined that they are the same, TYPE Without setting, the macro block outputs only the information that it is the same as the reference macro block, and does not perform the next INTRA processing or INTER processing. Upon completion of the INTRA or INTER processing, the target macroblock writes information specific to the encoding target macroblock as a macroblock header (709). If there is valid information (Coef) in a block in the macroblock, it is encoded (710) and written to the code buffer (711). This operation is performed for all N horizontal, vertical, and vertical macroblocks (712, 713), and the picture processing Pic P (Num) ends.

FIG. 8 shows a detailed processing flowchart of the INTRA processing 7.04 and the INTERA processing 705 of FIG.

In both the INTRA process and the INTERA process, the macroblock to be encoded is first decomposed into blocks (801, 802). In the case of INTRA, the block is DCT (803) and converted to DCT coefficients (Coef) (803). Next, the DCT coefficient is quantized to obtain QCoef (805). Part of Q Coef is encoded. The other part of QCoef is inversely quantized to IQCoe; f (807), and further DCT inversely transformed to RB1Ock data (809). The RB1Ock data is directly decoded in the INTRA processing (811).

In the case of the INTER processing, the difference between the block in the closest reference macroblock and the difference is calculated and used as B 1 ock data (820). The data is DCT sequentially as in the INTRA process (804), quantized (806), QC oef, and encoded. Also, Q oef is inverse quantization (8 06), inverse DCT (

808), and is added to the block from which the difference has been obtained first, and is added to be decoded block data (810).

FIG. 9 shows the encoding information of the video data encoded by the above processing. At the head is a sequence header 91. In the sequence header 91, mode information 911 such as a video capture cycle time, information 912 related to the entire captured instantaneous moving image such as an image size, and other information 913 are recorded. I have. Next, the picture header 92 is recorded. The picture header 92 contains time information 921 of the picture (image frame) and the type of the image frame.

Necessary information is recorded in image frame units of 9 22 and other 9 23. Next, Picture de 93 is recorded. In the picture data 93, a macro block header 931 is recorded for each macro block. The macroblock header 931 records information required for each macroblock, such as the type II motion vector of the macroblock and the quantization coefficient. However, the information in the macro block header 931 differs depending on the type of macro block. For example, in the case of the macroblock type IINTRA processing, the motion vector information is not included.

Following the macroblock header 931, the macroblock data 932 is recorded. Macro block data 932 contains the data obtained by performing DCT on the blocks in the macro block and quantizing them. In other words, blocks that have no data to encode as a result of quantization It is possible that there exist That is, the data of all the blocks in the macro block is not always included in the macro block data. The picture header 92 and the picture data 93 are converted from the reference image frame Pic (0) to the image frames Pic (—4), Pic (−3), Pic (−2), and Pic (−2). -1), Pic (4), Pic (1), Pic (2), Pic (3) are recorded as one image of the instantaneous moving image captured by the digital camera. Become.

10 and 11 are diagrams showing the configuration of another embodiment of the digital camera according to the present invention. In this figure, the point different from the embodiment of FIG. 1 is a portion surrounded by a dotted line. The same functions and components as those in FIG. 1 are denoted by the same reference numerals, and detailed description is omitted. In this embodiment, as shown in FIG. 11 in which the portion surrounded by the dotted line in FIG. 10 is shown in more detail, the capacity of the frame memory 22 of the input unit is set to M3 and M4 for two image frames. I have to. On the other hand, the code buffer 12 has memories 04, 05, '09, which store the compressed image frame data for six image frames, and input and output switching switches 02, 0 3 and a memory control unit 0 1 that performs the switching control. That is, in the present embodiment, by compressing and encoding the image frame before the one-time channel and storing it, the memory capacity is smaller than in the first embodiment in which the image frame data is stored as it is. Can be greatly reduced FIG. 12 shows a recording video frame and its compression method according to the second embodiment. The range of the recorded video is the same as in the first embodiment (FIG. 5). As for the compression method, the image frames before the shutter chance are encoded in the frame, and the encoding after the shutter chance is predicted from the shutter frame image frame or the shutter start image frame. Prediction (inter-frame coding) from the compressed image frame is performed and compressed.

FIG. 13 is a processing time chart in which the second embodiment of FIG. 10 performs compression encoding. The input image frame shot at the video capture period set in the mode input section 6 is used when the release signal from the shutter release button 1 and the shutter release button 5 are active (low level in the figure). Import to frame memory 22. In the figure, the third image frame is loaded into frame memory FM 1 and the fourth image frame is loaded into frame memory FM 2, and at the same time, the third image frame is read from frame memory FM 1. Intra-frame encoding and writing to code buffer 04 (B1). The fifth image frame is written to the frame memory FM 1 where the third image frame was stored. At the same time, the fourth image frame is read out from the frame memory FM2, encoded in the frame, and written into the code buffer 05 (B2). Similarly, the sixth, seventh and eighth image frames are processed as shown in the figure. The result of intra-frame encoding of the ninth image frame is overwritten by the code buffer 04 (B1). In this figure, the shutter was pressed at the 10th image frame. Shows the case. 11 At the same time as loading the 1st image frame into the frame memory FMl, read out the 10th image from the frame memory FM2, encode it in the frame, and compress it into the code buffer 05 (B2). Write out the sign. At the same time, the 10th image frame (the 10th decoded image frame) is written to the frame memory B9. 11 At the same time as loading the image frame No. 2 into the frame memory FM2, reading the image frame No. 11 from the frame memory FM1 and reading the image frame No. 11 from the image frame No. 10 stored in the frame memory B9. Perform predictive coding (interframe coding). The result is stored in the code buffer B3. At the same time, the compressed code No. 10 is read from the code buffer B2 and stored in the storage device. Also, the 11th image frame is stored in the frame memory B9. 13 Simultaneously load the image frame No. 3 into the frame memory FM1, read the image frame No. 12 from the frame memory FM2, and read the image No. 11 stored in the frame memory B9. Predictive coding (inter-frame coding) is performed from the frame. The result is written to code buffer B2. At the same time, the compression code of the sixth image frame stored in the code buffer B4 is read out and stored in the storage device. Also, the 12th image frame is stored in the portion of the frame memory B9 where the 10th image frame was stored.

14 When the image frame No. 4 is loaded into the frame memory FM2, the image frames No. 13 to 13 are read from the frame memory FM1 and the image No. 12 stored in the frame memory B9. Fret -Predictive coding (interframe coding) is performed from the system. The result is stored in the sign buffer B4. At the same time, the compressed code of the seventh image frame stored in the code buffer B5 is read out and stored in the storage device. The 13th image frame is stored in the portion of the frame memory B 9 where the 11th image frame is stored. Next, the 14th image frame is read from FM2, and prediction coding (interframe coding) is performed from the 13th image frame stored in the frame memory B9. Stored in buffer B5. At the same time, the compression code of the eighth image frame stored in the code buffer B 6 is read out and stored in the storage device. Next, the compression code of the ninth image frame stored in the code buffer B1 is stored in the storage device. Subsequently, the compression codes of the image frames of No. 11 of B3, No. 12 of B2, No. 13 of B4 and No. 14 of B5 are stored in the storage device.

In the embodiment shown in FIGS. 1 and 10, in order to briefly show the principle of the present invention and its operation, the capacity of the frame memory, the capacity of the code buffer, the operation timing, etc. were optimized. The configuration is not shown. By making the operation timing finer than the operation of the input image frame unit, the capacity of the frame memory and the code buffer can be reduced. Further, the frame memory B 9 in the compression encoder 23 operates to store the decoded image frame as a reference image frame. However, in the case of an instantaneous moving image, since the accumulation of errors during prediction is small, the decoded image frame is Instead, the input image frame can be used as it is as a reference image frame for prediction. Also, in the above description, by changing the setting of the mode switch of the apparatus shown in FIG. 1 or FIG. 10 using the image frame of the one-shot shutter as the reference image frame, The image frame of the one-shot event and the reference image frame may be different. Fig. 14 shows an example where the image frame of the evening-chance is set to 10 and the reference image frame is set to 8 or 12.

FIG. 15 is a diagram showing a configuration of a digital camera according to a third embodiment of the present invention. The optical signal of the subject captured from the lens 26 of the main body of the digital camera 25 is converted into an electrical signal by an image sensor 27 such as a CCD, and is input to a video interface 28. The video interface 28 temporarily stores a video signal in the memory 31 in the memory 31 in accordance with a signal from the shutter release button 29 and the mode button 30. Also, the input video is displayed on the display 32. The video interface 28 is connected to the control bus 34, address bus 35, and data bus 36 of the processor 33, and inputs input video data to the processor 33. The processor 33 receives the image frame stored in the memory 31 from the memory 31 by a control signal output from the shutter release button 29 to the control bus 34 through the video interface 28. Read the appropriate image frame and compress and encode it. At this time, memory 37 is used as the work memory and code buffer required. It is. The compression code stored in the memory 37 is read out by the storage controller 38 and recorded on the recording medium of the storage device 39. Further, each of the buses 3 4 3 6 is connected to an external port 41 via an external interface 40. The external port 41 is used, for example, to capture a captured image into a device such as a personal computer, and the storage control device 38 reads out the compression code stored in the storage device 39, and the external interface 4 Operates to transfer the data to the computer through 0.

The captured video can also be played back with a digital camera 25. The storage device controller 39 receives the reproduction request from the processor 33, reads the appropriate compression code, and decodes the compressed code recorded in the storage device 39 using the memory 37. (Decompression) The decoded image frame is written to the memory 31 through the video interface 28. The image interface 26 operates to read out the image frame as appropriate and display it on the display 32. The digital camera device 25 includes a power source 42 such as a battery and a battery, and a clock generator 43 that generates an overall basic clock. The power supply 42 can store power from the port 44, and this device also has a portable function.

Industrial applicability

According to the present invention, (1) by recording a subject as an instantaneous moving image, it is possible to record movement and changes in facial expressions. (2) Shutdown To record the video before and after one point, the image of the part to be recorded Can be reliably photographed. (3) By composing a single instantaneous image with a reference image frame and a plurality of image frames predicted from the image frame, compression is achieved compared to the conventional continuous shooting operation of a still image camera. A device that captures high-quality video with improved efficiency is realized. In addition, by recording the reference image frame at the beginning of the compressed code string, it has become possible to easily detect and edit necessary images.

Claims

The scope of the claims

1. a first step of converting the video signal into a series of digital image frame signals;

A second step of selecting a signal of one of the consecutive image frames as a signal reference image frame, and a third step of intra-frame encoding the signal of the reference image frame. ,

The signals of the image frames temporally before and after the reference image frame of the plurality of image frames are encoded by inter-frame coding from the reference image frame or from the reference image frame to the frame. A fourth step of performing inter-frame coding from the inter-coded image frame;

A fifth step of storing the signals encoded in the third and fourth steps in a storage means.

2. In the fifth step, the intra-frame coded signal obtained in the third step is recorded at the beginning of the storage in the storage means, and then the interframe code obtained in the fourth step is recorded. 2. The video imaging / accumulating method according to claim 1, wherein the encoded signal is recorded.

3. A first means of capturing video and capturing a series of digital image frames;

Selecting means for selecting one of the series of digital image frames as a reference image frame;

Intra-frame encoding of the signal of the reference image frame The signals of the image frames before and after the quasi-image frame are interframe-coded from the reference image frame or from the inter-frame-coded image frame from the reference image frame. A compression encoding unit that performs inter-frame encoding of

A video imaging device having a storage unit for recording the compression-coded signal as an instantaneous moving image.

4. The storage unit in which the storage unit stores the coded signal coded by the compression coding unit as a bit stream, and the intra-frame coded signal at the head of the bit stream. 4. The video imaging device according to claim 3, further comprising storage control means for controlling the recording of the inter-frame coded signal following the above.

5. The video imaging device according to claim 3 or 4, wherein a means for decoding the compression code recorded in the storage unit and reproducing the image is added.

6. The first frame memory in which the selection means stores the plurality of image frames, a release button, a shutter button, and an image frame to be input while the release button is pressed. The frames are sequentially stored in the first frame memory, and the signal of the image frame of the shutter channel whose shutter button is pressed is read out from the first frame memory to the compression encoding unit. 5. The video imaging apparatus according to claim 3, further comprising: a reading control unit that reads a signal of an image frame before and after the shutter chance and reads the signal to the compression encoding unit according to the signal.

7. The video imaging device according to claim 6, further comprising means for decoding a compression code recorded in the storage unit and reproducing an image.

8. The selecting means selects a release button for capturing the plurality of image frames, and one of the captured image frames, and outputs the selected frame as the reference image frame to the compression encoding unit. The code buffer according to claim 3 or 4, further comprising a code buffer for storing a signal of the compression-encoded vibration between the compression-encoding unit and the storage unit. Video imaging device.

9. The video imaging device according to claim 8, further comprising means for decoding a compression code recorded in the storage unit and reproducing an image.