JP2807791B2 - Image signal recording device - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION [Object of the Invention] (Industrial application field) The present invention relates to an image signal recording apparatus in which transfer of an image signal from an image sensor is improved in connection with encoding.

(Prior Art) An optical image is converted into an image signal and this image signal is stored as digital data in a storage device. For example, in a digital electronic still camera, an image signal having a huge data amount is stored in a storage device having a limited capacity. Therefore, information compression is indispensable.

Here, a schematic configuration of the digital electronic still camera will be described with reference to FIG.

First, an optical image incident through a lens system and a shutter (not shown) is photoelectrically converted by a solid-state imaging device 11 to become an image signal for one screen.
After that, it is converted into a digital signal by the A / D converter 13. This digital signal is also encoded by an encoder 14 for information compression.
After the data amount is compressed, the data amount is stored in the image storage device 15. Thus, with a digital electronic still camera,
Compression of image information by the encoder 14 is indispensable.

Conventionally, various methods have been studied as image information compression methods for transmitting a still image. However, the purpose of information compression in image transmission is mainly to reduce the bit rate on a transmission path, and is not suitable for downsizing hardware. Then, as described in JP-A-62-176227, for example, the A / D-converted image information is temporarily stored in a buffer memory such as a field memory or a frame memory, and the correlation of the stored information is checked. After that, encoding is generally performed using an optimal encoding method.

However, in such a method, a buffer memory having a storage capacity several stages larger than that of a memory for one screen for storing image information after compression must be used as the buffer memory described above.
For this reason, miniaturization and cost reduction strongly demanded as an electronic still camera cannot be satisfied.

Further, in order to enable high-speed continuous shooting which is a feature of the still camera, encoding must be performed in almost real time. However, in the configuration described in JP-A-62-176227, an image must be divided into blocks and a block distortion signal must be calculated for one entire field.
The encoding is not completed until the calculation of the entire field is completed. In other words, it takes time to calculate at least one field until a signal is transmitted. Also,
Since many calculations need to be processed in a short time, there is a problem in downsizing the hardware.

As a method to solve such a problem, an image information compression method using DPCM has been previously announced (IEICE, February 26, 1988, Technical Research Report, "Image Information Compression Using DPCM"). Consideration of method "). In this method, the block size is determined in advance so that the deviation of the coded signal from the original signal is minimized, and the coding is performed by non-linear conversion adapted to the dynamic range of each block. While reducing time, and
Encoding can be performed without the need for a field memory for encoding. That is, in the present method, a prediction error signal obtained by subtracting a predicted value (a value obtained by predicting a current value from a previous sample point) from a current sample value is nonlinearly converted and transferred to a memory.

 Hereinafter, an outline of this method will be described with reference to FIG.

First, an image signal for one screen imaged by an image sensor (not shown) is converted into a digital signal by an A / D converter 13, and then n × n by a block divider 17 using local correlation of the image. The block is divided into a plurality of square blocks each composed of a plurality of pixels, and is stored in the n × n block memories 18.

Prior to encoding, the predictor 20 and the maximum prediction error detector 21 check the maximum value or standard deviation of the absolute value of the prediction error signal in each block, and thereby select the optimal nonlinear conversion function for each block. To this end, switching control is performed on switching units 23 provided corresponding to the plurality of nonlinear converters 22. Then, a plurality of nonlinear converters 22 are provided corresponding to, for example, the maximum value or the standard deviation of the absolute value of the prediction error signal.

Further, a local decoder 25 is provided to obtain the above-described predicted value. The local decoder 25 includes the same number of nonlinear inverse converters 26 provided corresponding to the plurality of nonlinear converters 22, and the nonlinear converters 22. Switch which is provided to add the output of the non-linear inverter 26 to one of the non-linear inverters 26, and which is switched and controlled in conjunction with the switch 23.
27 and a predictor 28 are configured as shown. Then, the output of the local decoder 25 is added as a prediction value for the current sample value output from the block memory 18 to obtain these differences. Then, the prediction error signal, which is the difference between them, is encoded by the optimal nonlinear converter 22 selected by the switch 23 and stored in the image storage device 15.

And, when the encoding method is used, as in the conventional example,
A buffer memory for encoding does not need to have a large storage capacity of one field or one frame memory, and only the n × n block memories 18 need be used.

However, since scanning is performed with a normal television signal, the calculation of the budget error signal for the first n × n block does not end unless the data of the nth line is A / D converted as shown in FIG. . Then, if the next (n + 1) th line of data is input before the encoding is completed by the nonlinear converter 22 selected based on the result, it cannot be stored in the memory.

One way to solve this is to use an n-line memory,
That is, it is conceivable to provide two memories for storing all data up to n lines, input data alternately, and use one of the memories as a buffer for data of the next n lines. That is, this is a method in which data stored in one n-line memory and data from n + 1 to 2n lines input during encoding of a plurality of blocks are stored in the other n-line memory.

However, for this purpose, the encoding of the data for n lines must be completed within the scanning time for n lines, which is difficult to realize due to factors such as the calculation speed.

(Problem to be Solved by the Invention) As described above, reducing the size of the buffer memory for encoding has a practical problem and is difficult to realize.

The present invention provides an image signal recording apparatus that can significantly reduce the capacity of a buffer memory for encoding without affecting the time required for encoding by improving the method of extracting image signals from the image sensor. The purpose is to provide.

[Configuration of the invention]

(Means for Solving the Problems) An image signal recording apparatus according to the present invention converts an optical image into an image signal for one screen composed of a large number of pixel data and holds the pixel data. Transfer control means for extracting and transferring pixel data for a predetermined section of the image signal held in the element, an A / D converter for converting the pixel data for the predetermined section extracted by the transfer control means into a digital signal, and A buffer memory for storing the converted digital signals of a predetermined section, an encoder for encoding the digital signals stored in the buffer memory for each predetermined block, and digital data encoded by the encoder. A transfer device that changes a transfer rate, which is a transfer rate of pixel data, and stores the digital data stored in the buffer memory. The pixel data of the next predetermined section is not transferred until the encoding of all the data is completed.

(Function) The present invention extracts the image signal for one screen held by the image sensor and the pixel data for a predetermined section, transfers the pixel data at an arbitrary speed, and transfers the pixel data for the next predetermined section. In order to prevent the encoding from being performed until all the encoding of the data previously extracted and stored in the buffer memory for each block is completed, without affecting the encoding speed,
The buffer memory for encoding can be significantly reduced in capacity.

Hereinafter, an embodiment of the present invention will be described with reference to the drawings.

In FIG. 1, the solid-state imaging device 11 has a lens
An optical image enters through the aperture 30, the diaphragm 31, and the shutter 32.
The solid-state imaging device 11 can convert an optical image into an image signal for one screen including a large number of pixel data, and can hold each of the pixel data. That is, as shown in FIG. 2, the solid-state imaging device 11 includes a large number of photoelectric conversion units, such as photodiodes, which are regularly arranged and serve as pixels.
34 and multiple vertical charge transfer devices, so-called vertical CCD3
5 and a horizontal charge transfer element common to them, a so-called horizontal CCD 36.

Then, the solid-state imaging device 11 photoelectrically converts the optical image,
At this time, electric charges as pixel data generated by the respective photodiodes 34 are simultaneously transferred to the vertical CCD 35. At the television rate, this read pulse is located in the vertical blanking period. Thereafter, the vertical CCDs 35 transfer the charges one after another at the horizontal synchronization frequency, and the horizontal CCD 36 transfers the charges every one horizontal period, and this becomes an output signal. As described above, the pixel data held in the solid-state imaging device 11 is sequentially transferred for each horizontal line.

Drive pulses for the vertical CCD 35 and the horizontal CCD 36 are supplied from a drive pulse generation circuit 37 in FIG. Therefore, this drive pulse generation circuit
The transfer control means 40 having a variable transfer rate, which is a transfer rate, is constituted by the 37, the synchronizing signal generating section 38, and a part of the function of the system control section 39 for controlling them.

The pixel data transferred from the solid-state imaging device 11 is
After passing through the signal processing circuit 12 as before, the A / D converter 13
, And is stored in the encoding buffer memory 41. Further, the encoder 14 encodes the digital signal stored in the buffer memory 41 for each block of n × n pixels to compress the information amount.
This is the same as the configuration following the block memory 18 shown in FIG. The image signal for one screen, which is encoded by the encoder 14 and the information amount is compressed, is stored in the image storage device 15. Further, the system control unit 39 controls the signal processing circuit 12, the buffer memory 41, the encoder 14, and the image storage device 15, including the transfer control means 40, in an integrated manner.

Here, the charges generated in the photodiodes 34 on the solid-state imaging device 11 are generally immediately transferred to the output terminal of the solid-state imaging device 11 by the CCDs 35 and 36, respectively. For example, in the case of a television rate, the charge photoelectrically converted in 1/60 seconds is
It is normal for everything to be transferred in 60 seconds.

On the other hand, in the present embodiment, the transfer of the charges, that is, the pixel data, is controlled by the transfer control means 40 as follows. That is, all pixel data of a predetermined section of the image signal held in the solid-state imaging device 11, for example, all the pixel data of the section from line 1 to line n in FIG.
And the horizontal CCD 36 sequentially transfers the signals at normal timing. After the transfer to the n-th line, the drive pulse is temporarily stopped and the transfer is not performed.

The transferred pixel data is subjected to signal processing, converted into a digital signal, and stored in the buffer memory 41.
The buffer memory 41 only needs to store the pixel data up to the n-th line shown in FIG. 5, and the memory capacity is much smaller than that of the conventional one, and the buffer memory 41 is an n-line memory.

The digital signal stored in the n-line memory 41 is
The image data is encoded for each of a plurality of blocks of n × n pixels and stored in the image storage device 15.

Then, when the encoding for each block is completed, the transfer control means 40 gives a drive pulse again to the vertical CCD 35 and the horizontal CCD 36, and reads out the pixel data from the next (n + 1) th line to the 2nth line in FIG. The same processing is repeated. In this manner, the image signals are sequentially encoded, and one screen is stored in the image storage device 15.

Next, this operation will be described with reference to the flowchart of FIG.

First, from each photodiode 34 shown in FIG.
The charge is transferred to the CCD 35 (step 1). Thereafter, charge transfer is performed once to the vertical CCD 35 (step 2), and then charge transfer is performed to the horizontal CCD 36 (step 3). The transferred charge is signal processed, sampled and held (step 4), A / D converted (step 5), and stored in the buffer memory 41 for one line (step 6).

Next, it is determined whether the charge transfer of the vertical CCD 35 has been completed n times (step 7).
Repeat until On the other hand, if the processing is completed n times and the pixel data of n lines is stored in the buffer memory 41, it is divided into P blocks of n × n pixels (step 8). Then, encoding is performed for each of the divided blocks (step 9) and stored in the image storage device 15 (step 10). Next, it is determined whether the encoding has been completed for all the P blocks (step 11), and if not, steps 9 and 10 are repeated. On the other hand, if all the coding has been completed, it is determined whether the coding for one screen has been completed (step 12).
And the transfer operation of the vertical CCD 35 is performed again to transfer the charge on the (n + 1) th line. Thereafter, the same operation is repeated, and END is obtained when the encoding of one screen is completed.

In the above embodiment, the charge transfer rates of the vertical CCD 35 and the horizontal CCD 36 are set to the TV rate. In this case, the time required for encoding is a value close to the real time because only the nonlinear converter 22 shown in FIG. 4 is selected and data conversion is performed by this. Of course, transfer at a rate faster than the TV rate, and if the sum of this transfer time and the transfer pause time is equal to the CCD transfer time for n lines at the normal TV rate,
The rate from the encoding to the storage in the image storage device 15 is exactly equal to the television rate, realizing real-time encoding.

In any of the above cases, high-speed continuous shooting which is a feature of the still camera can be sufficiently coped with.

Further, when the processing time of the encoding processing circuit at the subsequent stage is slow, by setting the charge transfer rate lower than the normal television rate in accordance with the processing time, it is possible to reduce the capacity of the buffer memory in the same manner.

Note that a solid-state imaging device 11 that can randomly read arbitrary pixel data, such as a MOS type, is used, and if a mechanical shutter mechanism is further provided, the light after the exposure time is blocked. Since the photoelectric conversion unit (eg, photodiode) of each pixel can be used as a short-time analog memory, the data amount of a predetermined section transferred at one time is not n lines but n × n pixels. be able to. Therefore, as the buffer memory 41 for encoding, n × n pixels (one block)
With this memory capacity, encoding is possible, and the capacity of the buffer memory 41 can be further reduced.

The block division is not limited to the square of n × n, but may be of course n × m pixels.

〔The invention's effect〕

According to the image signal recording apparatus of the present invention, when one screen is divided into blocks and encoding is performed for each block,
The buffer memory required for encoding can be reduced in capacity without affecting the encoding speed, and the hardware can be reduced in size.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of an image signal recording apparatus according to the present invention, FIG. 2 is an explanatory view showing a configuration example of the image pickup device shown in FIG. 1, and FIG. 3 is shown in FIG. FIG. 4 is a block diagram for explaining the encoder, FIG. 5 is an explanatory diagram for explaining the relationship between pixel data on an image sensor and blocks, and FIG. 6 is a conventional example. FIG. 11 solid-state imaging device, 13 A / D converter, 14 encoder, 15 image storage device, 40 transfer control means, 41
Buffer memory.

──────────────────────────────────────────────────続 き Continued on the front page (58) Field surveyed (Int.Cl. ⁶ , DB name) H04N 5/91-5/956 H04N 7/24

Claims (1)

(57) [Claims] An image sensor capable of converting an optical image into an image signal for one screen composed of a large number of pixel data and holding the pixel data, and an image signal held by the image sensor for a predetermined section. Transfer control means for extracting and transferring pixel data; an A / D converter for converting pixel data of a predetermined section extracted by the transfer control means into a digital signal; and storing the converted digital signal of the predetermined section. A buffer memory; and an encoder that encodes the digital signal stored in the buffer memory for each predetermined block, and a storage device that stores digital data encoded by the encoder. The transfer rate, which is the transfer rate of the pixel data, is variable, and the following steps are performed until the encoding of all the digital data stored in the buffer memory is completed. An image signal recording apparatus configured to not transfer pixel data for a fixed section.