JPH0686207A - Digital still camera - Google Patents

Abstract

PURPOSE:To expand the dynamic range of a digital still camera by compressing a white part and also a black part by applying prescribed nonlinear quantization. CONSTITUTION:Video signal output of R, G, and B from a CCD 8 are amplified by a pre-amplifier 9, and are supplied to an ADC 11 at the next stage, then, they are digital-converted. Color video signal data of 10 bits from the ADC 11 is supplied to an image processing circuit 12, and R, G, and B signals are supported with a Y/C converter 4, and after they are converted to a luminance signal Y and a color difference signal C, they are supplied to a nonlinear converter 13. The circuit 13 is comprised of a white compression part 13a and a black compression part 13b that is an ordinary gamma correction circuit part, and the white part and the black part of a video are compressed at the circuit, and input data of 10 bits is outputted after being converted to the one of eight bits. which generates a bit conversion effect. In other words, the compression by nonlinear quantization can be performed on the black part and the white part based on the control of every kind of data from the circuits 12, 13 and a CPU 17 via a bus 21 and nonlinear quantization data from a ROM table.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a digital still camera for recording still images and, more particularly, to improvement of an image quantizing circuit.

[0002]

2. Description of the Related Art As a conventional digital still camera, for example, a technique disclosed in Japanese Patent Laid-Open No. 3-104486 is known.

In the above publication, when the difference in light amount between the bright portion and the dark portion of the subject is small, if the gradation conversion by the standard γ (gamma) value is performed, the contrast ratio is weak, so that it is visually recognized. A good image cannot be obtained. On the contrary, in order to solve the problem that when the light amount difference between the bright part and the dark part of the subject is large, the contrast becomes weak by the γ correction with the standard γ value, and similarly a good image cannot be obtained. It is preferable that the contrast is weakened by the gradation converting means when the gradation conversion of the image signal is performed when the light amount difference between the respective parts of the subject in the image pickup screen is large, and conversely when the light amount difference is small, the contrast is increased. It is shown that the image is obtained.

That is, specifically, a plurality of areas of the subject are measured, and one of a plurality of gradation conversion means that can be selected in accordance with each light amount measured is selected, so that gradation is achieved. As the conversion characteristic, various γ correction characteristics are used.

[0005]

In the digital still camera having the above-mentioned conventional configuration, the degree of the γ characteristic is stepwise switched by a switch or the like. Partial compression is performed, but this is N
It comes from the relationship of the monitor on the television receiver side in the TSC system, and it is not inevitable with the image data.

The image pickup signal is converted into digital data through an analog-digital conversion circuit (ADC) before the compression of the white portion by γ correction, but normally, the ADC having about 8 bits is used. The gradation expression in 8 bits is 256 gradations. Black is now 0 and white is 100%
Then, only about 200 gradations can be obtained.

On the other hand, looking at the relationship between the discrimination noise level and the lightness against the noise interference of the image data, as shown in FIG. 5, human eyes have a weak noise discrimination power in the black part 1 and the white part 3, but in the middle gray part 2. It is known that the noise discrimination power is strong in the range.

Therefore, the present invention intends to effectively use the redundant portion of the black portion by truncating the black portion having a weak noise discrimination power by human eyes and performing the non-linear quantization in the black portion simultaneously with the bit compression.

[0009]

In the digital still camera of the present invention, a video signal representing a still image is stored in the storage means 15 as digital data, as shown in FIG. In the digital still camera 10, a non-linear quantizing means 13a for compressing a standard white part
And a non-linear quantizing means 13b for compressing the black portion.

[0010]

According to the digital still camera of the present invention, it is possible to expand the dynamic range by compressing the white portion as well as the black portion by the predetermined non-linear quantization.

[0011]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An embodiment of the digital still camera of the present invention will be described below with reference to the drawings.

FIG. 1 shows only one system path of red (R) or the like among the color image pickup means of the three-plate CCD. Imager 4
Digital still camera 10 with connector 1
4 and a memory 15 which is detachably connected to the memory 15. In addition, the camera 10 may have a buffer memory inside. Further, instead of the semiconductor memory, another storage system such as magnetic recording may be used. In the figure, the element portion above the connector 14 is mounted as a digital still camera 10 in a single housing.

The memory 15 is a rewritable digital storage device in which, for example, an SRAM semiconductor memory is mounted in the form of a "module" such as an integrated circuit (IC) card or a cartridge, and data input / output lines, addresses, and reading are performed. Control lines including / write enable, chip select, strobe, clock, etc. are connected to the camera 10 via the connector 14. In the connector 14 and the memory 15, for example, when an image of one frame is represented by data of 1 M to 1.5 M bits, an SRAM having a storage capacity of 16 M bits per chip can realize a storage device for taking 24 frames with two chips. .

The image pickup section 12, as shown in the drawing, captures still images of the image pickup lens 5, diaphragm 6, shutter 7, image pickup device 8, photometric / distance measuring mechanism, viewfinder (not shown) and their driving mechanism. The focusing of the imaging lens 5, the control of the diaphragm 6, the opening and closing of the shutter 7, etc. are controlled by the CPU 17 via the bus line 20. The image pickup device 8 is a solid-state image pickup device such as a CCD or a MOS. A color filter 8a is attached to the image pickup cell array of the color solid-state image pickup device shown in FIG. The video signal color-modulated in response to the clock from 19 is sequentially output to the output in a point (pixel) manner. An appropriate color segment array of the color filter 8a is used. In the case of the 3 type plate, R, G and B color filters are arranged in front of the CCD.

The R, G and B video signal outputs from the CCD 8 are pre-amplified by a pre-amplifier 9 and supplied to an ADC 11 in the next stage to be digitally converted. This ADC 11 is AD with 10 bits in this example. The synchronizing signal from the synchronizing signal generating circuit 16 is supplied to the ADC 11, the image processing circuit 12 and the non-linear converting circuit 13 which will be described later.

The 10-bit color video signal data from the ADC 11 is supplied to the image processing circuit 12 so that R, G, B
Y for converting the signal into Y (luminance signal) and C (color difference signal)
Image processing such as / C conversion, 4: 2 · 2 conversion, or knee processing is digitally performed and then supplied to the non-linear conversion circuit 13.

The non-linear conversion circuit includes a white part compression part 13a and a black part compression part 13b which are normal γ correction circuit parts.

As will be described later, the non-linear conversion circuit 13 compresses a white part and a black part of an image, and the 10-bit input data is converted into 8-bit data and output, so that a bit conversion effect is also provided. You can also do it. The image processing circuit 12 and the non-linear conversion circuit 13 control various data from the CPU 17 via the bus 21, and perform non-linear quantization compression for the black and white parts based on the non-linear quantized data from the ROM table.

The image data compressed by the non-linear conversion circuit 13 and subjected to γ correction is stored in the IC card memory 15 via the card memory interface 18.

Next, one embodiment of the non-linear conversion circuit 13 will be described in detail with reference to FIGS.

FIG. 2A shows the input / output characteristic of the γ correction showing the normal gradation characteristic. When the output from the image processing circuit 12 on the horizontal axis is x and the output from the nonlinear conversion circuit 13 on the vertical axis is y, the γ characteristic is generally expressed by the following equation (1). Here, the larger the value of γ, the stronger the contrast becomes, and in the case of displaying a still image through a monitor in the reproduction system, γ = 0.
Select about 7.

Here, what is nonlinearly quantized by γ is a portion having a large value on the x-axis, that is, an input data portion corresponding to the white portion 3 of the image, and the black portion 1 is directly quantized. ing. For this reason, if the black part 1 of the image has 256 gradations of 8 bits as described above, the ratio of the total gradation of the black part is large, and the actual gradation is about 200 gradations. As described above, since the black portion has a low noise discrimination power when the image is observed by human eyes, the noise level is not noticeable even if the black portion 1 is nonlinearly quantized and compressed. It becomes possible to adjust.

That is, the black portion 1 of the straight line portion in FIG.
When nonlinearly quantized and compressed like the black part 1'of the gray part 2 '
It is possible to expand the gradation expression of the white part 3 '.

Actually, the black portion 1'allows the level of 30 IRE or less to be compressed to 1/2 to 1/3. Of course, the compression can be continuous or linear.

For such non-linear quantization compression from the white part to the black part, a table such as a ROM is provided in the non-linear conversion circuit unit 13, and gradation conversion is performed based on this conversion table. An example of this configuration will be described with reference to FIG.

10 of R, G, B from the image processing circuit 12
The bit image data includes white and black compression units 13a and 1 respectively.
3b is supplied to the non-linear conversion circuit 13, and as shown in FIG.
RAM (Static Rondom Access Memory) 13RS, 1
It is supplied to 3GS and 13BS.

Further, in this example, the table data for the γ correction characteristic curve, which includes the white to black portions shown in FIG. 2B, is stored in EPROMs (Erasable and Programmable) for R, G, and B, respectively.
e ROM) 13RE, 13GE, 13BG. Since these EPROMs have a slow reading speed, the EPROM 1 is read when the power of the switch SW shown in FIG.
Outputs 8-bit stored data of 3RE, 13GE, and 12BE.

The SRAMs 13RS, 13BS, 13GS as well as the EPROMs 13RE, 13GS, 13BE have a CPU
Clock and address data are supplied from the memory 17 and the transfer address counter 22, and the CPU 17 transfers 10-bit image data to the SRAM 13RS, 13GS, 13BS.
Write to the specified address of SRAM, 13RS, 13 of SRAM
Outputs 8 bits of GS and 13BS. The CPU 17 performs writing, reading, γ selection, and the like.

FIG. 4 is a system diagram showing an embodiment of the reproducing system of this example. In FIG. 15, reference numeral 15 denotes an IC card memory in which still image data captured by the digital still camera 10 is also compressed and stored in the black portion by the non-linear conversion circuit 13.
The plug 14 of the IC card memory 15 is inserted into the jack of the card interface 25 on the reproducing device side. The output of the card interface 25 is supplied to the movable contact piece a of the switching means 26. The CPU 35 displays the reproduced still image as a normal C image.
The movable contact piece a of the switching means 26 is displayed on the monitor 34 such as an RT or by printing a reproduced still image.
Is selected as the fixed contact b or c.

The fixed contacts b and c of the switching means 26 have first and second non-linear inverse conversion circuits 27 and 28, first and second image processing circuits 29 and 30, and first and second.
Digital-analog converter circuit (hereinafter referred to as DAC) 3
The output of the first DAC 31 is sent to the printer 33 via the 1 and 32, and the output of the second DAC 32 is sent to the monitor 34 such as a CRT.
And the hard print of the still image or the still image is displayed on the CRT, respectively.

In the present example, the first non-linear inverse conversion circuit performs expansion opposite to the compression converted by the non-linear conversion circuit 13 and converts the conversion characteristic into substantially linear to convert it into an analog signal after image processing and print it out. .

In the second non-linear inverse conversion circuit 28, γ = 0.
The correction is performed at 7, and a still image is displayed on the monitor 34 such as a CRT after image processing and analog signal conversion.

Since the digital still camera of the present invention is constructed as described above, since the black portion having a low lightness dependency of noise interference is compressed, the dynamic range is expanded as a whole, and the white portion of the image when the contrast is high is expanded. An image with good reproducibility and no visual unnaturalness can be obtained.
Further, even if a 10-bit ADC is used, the nonlinear conversion circuit 13 can perform bit conversion for 8-bit conversion.

[0034]

According to the digital still camera of the present invention, as a result, the dynamic range is expanded and an image free from unnaturalness can be easily obtained.

[Brief description of drawings]

FIG. 1 is a system diagram showing an embodiment of a digital still camera of the present invention.

FIG. 2 is a diagram illustrating a γ characteristic used in the digital still camera of the present invention.

FIG. 3 is a non-linear compression processing circuit diagram used in the digital still camera of the present invention.

FIG. 4 is a system diagram showing an embodiment of a reproduction system of the digital still camera of the present invention.

FIG. 5 is an explanatory diagram showing a lightness dependence rate of conventional noise interference.

[Explanation of symbols]

 1 White part 2 Gray part 3 Black part 10 Digital still camera 13 Non-linear conversion circuit 13a White compression part 13b Black compression part 15 IC card memory 17 CPU

Claims (3)

[Claims] 1. In a digital still camera configured so that a video signal representing a still image is stored as digital data in a storage means, a non-linear quantizing means for compressing a standard white portion and a black portion are compressed. A digital still camera, comprising: 2. The digital still camera according to claim 1, wherein the non-linear quantized data by the non-linear quantized means of the black portion by the non-linear means is stored in a ROM table. 3. When reproducing data from the digital data storage means of a digital still camera configured to store the digital data in a removable storage means via a non-linear quantization means for compressing the black portion. 2. The digital still camera according to claim 1, further comprising a decompressing unit for performing a non-linear inverse conversion of the black portion, and further comprising a selecting unit for selecting a decompression degree of the decompressing unit.