JP2003037780A - Image data acquisition method and device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To reduce noise components contained in output image data of a CCD element. SOLUTION: A high-speed analog/digital converter 21 samples a feed-through part of an output signal obtained by imaging an object by a CCD element 10 and applies analog/digital conversion to the sampled signal to obtain a plurality of reference signals R1, R2, R3 and obtains a reference signal MR from the signals R1, R2, R3 a low-speed analog/digital converter 22 samples a data part of the output signal and applies analog/digital conversion to the result to obtain a data signal D, and a CDS(correlation double sample) processing means 30 applies the correlation double sampling processing based on the acquired reference signal MR and the acquired data signal D to acquire image data.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image data acquisition method and apparatus, and more particularly to an image data acquisition method and apparatus for performing correlated double sampling processing on an output signal of a CCD element to acquire image data. is there.

[0002]

2. Description of the Related Art Conventionally, in an apparatus for acquiring image data using a CCD element, a correlated double sampling process (hereinafter abbreviated as CDS process) is carried out in order to reduce noise mixed in the image data. ing. This CD
In the S process, the data portion of the output signal output from the CCD element (the portion in which the signal carrying the image picked up by the CCD element exists) is sampled once, and the data signal before the data portion or This is a process of acquiring image data based on the difference between the reference signal acquired by sampling the feedthrough part, which is a non-signal part, output once, and by this process, the image data is spread over the data signal and the reference signal. The noise components contained in the image data are reduced by canceling out the components of noise (hereinafter, referred to as common noise) having a relatively long cycle, that is, having substantially the same size and overlapping with each other, that is, having a longer cycle than the feedthrough portion and the data portion. .

[0003]

However, since the signal level of the feed-through section is lower than that of the data section, the influence of noise is large, and when the reference signal is acquired by sampling the feed-through section, If noise having a shorter cycle (hereinafter referred to as short cycle noise) is mixed in the feed through section, the acquired reference signal will deviate from the value indicating the common noise component. In particular, when sudden spike noise with a short cycle occurs in the feedthrough portion, if this spike noise is sampled, the obtained reference signal will have a value that is significantly different from the value indicating the common noise component. . Therefore, there is a problem that even if the CDS processing is performed using the reference signal obtained by sampling such short-period noise, the sufficient effect of reducing the noise of the image data cannot be obtained.

The present invention has been made in view of the above circumstances, and an object of the present invention is to provide an image data acquisition method and apparatus capable of further reducing noise contained in image data.

[0005]

According to the image data acquisition method of the present invention, an output signal obtained by imaging with a CCD element is sampled by a feed through section and A / D converted to obtain a reference signal, and a data section is obtained. In the image data acquisition method of acquiring image data by performing correlated double sampling processing based on the reference signal and the data signal, the reference signal is fed to the same feed. It is characterized in that the through portion is sampled a plurality of times, A / D converted, and acquired.

In the image data acquisition apparatus of the present invention, the CCD element and the output signal of the CCD element are sampled by the feed through section and A / D converted to output the reference signal, and the data section is sampled by the A / D section. An image data acquisition device comprising: A / D conversion means for converting and outputting a data signal; and CDS processing means for acquiring image data by performing correlated double sampling processing based on the reference signal and the data signal, The A / D conversion means is characterized in that the output signal of the CCD element is sampled a plurality of times by the same feed-through section and A / D converted to output a reference signal.

The A / D conversion means samples the output signal of the CCD element a plurality of times in the same feedthrough section, performs A / D conversion, and outputs a reference signal at a relatively high speed. Means and the first A / D conversion means for sampling once in the data section and A / D converting and outputting the data signal, that is, the second A / D conversion at a relatively low speed. It is desirable to be composed of D conversion means. Further, this A / D conversion means is C
The output signal of the CD element may be a high-speed A / D converter that performs A / D conversion by sampling the same feed-through section and the same data section a plurality of times.

The correlated double sampling processing is an arithmetic processing for obtaining a value of one reference signal from the values of a plurality of reference signals or obtaining a value of one data signal from the values of a plurality of data signals. And so on, and such an arithmetic process is hereinafter referred to as preprocessing of correlated double sampling processing, and one data signal value and one reference signal obtained by performing this preprocessing are obtained. The processing of canceling the common noise component based on the value of is called post processing of correlated double sampling processing.

[0009]

According to the image data acquiring method and apparatus of the present invention, when the correlated double sampling process is performed to acquire the image data, the reference signal is acquired by sampling the same feed-through section a plurality of times. Therefore, for example, in the preprocessing of the correlated double sampling processing, the values of the plurality of reference signals acquired are arithmetically averaged, or the value of the reference signal having a unique value among the values of the plurality of reference signals. Correlation 2 using the value of the reference signal obtained by removing
The image data can be acquired by performing the post-processing of the heavy sampling process, and the noise component included in the acquired image data can be more reliably reduced as compared with the related art. That is, as described above, the influence is dispersed by averaging the short-period noise components generated in the feed-through portion, or the short-period noise components are directly removed to obtain the correlation 2 based on the reference signal obtained.
By performing the post-processing of the heavy sampling process, the correlation double based on the reference signal acquired by only one sampling in the feed-through section without the influence of the short cycle noise is reduced as in the conventional case. It is possible to further reduce the short cycle noise component included in the acquired image data, as compared with the case where the sampling process is performed.

Further, the A / D conversion means samples the same feed-through section a plurality of times, performs A / D conversion and outputs a reference signal at high speed, and a sampling is performed in the data section. A / D conversion is performed at a lower speed than the first A / D conversion means for outputting a data signal.
If it is composed of a low-speed second A / D conversion means for performing D conversion, the feedthrough portion of the output signal has a small fluctuation range of the signal value, and the quantum of A / D conversion for obtaining a predetermined resolution. Since the number of digitizations (the number of bits to be A / D converted) can be reduced, high-speed A / D conversion that accompanies sampling multiple times can be implemented relatively easily.
On the other hand, in the data part where the fluctuation range of the signal value is large and the quantization number of A / D conversion (the number of bits to be A / D converted) needs to be increased in order to obtain a predetermined resolution equivalent to that of the feedthrough part. Even if there is, it only needs to be sampled once,
The A / D conversion of the data section can be performed at a lower speed than the A / D conversion of the feed through section. Therefore, the A / D in the feedthrough section and the data section
A / D with different speeds as described above, each dedicated to conversion
If the converting means is applied, the same high-speed A / D converting means can be used for the A / D in the feedthrough part and the data part
The device cost can be reduced as compared with the case where the conversion is performed (that is, the case where the high-speed A / D conversion with a large number of quantization is performed by one A / D conversion means).

When the A / D conversion means is a high-speed one that performs the A / D conversion by sampling the same feed-through section and the same data section a plurality of times, respectively, not only the feed-through section is obtained. With respect to the data part, the same effect of reducing the short-period noise component as described above can be obtained, and the noise component included in the acquired image data can be further reduced.

[0012]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Specific embodiments of the present invention will be described below with reference to the drawings. 1 is a block diagram showing a schematic configuration of an embodiment of an image data acquisition device of the present invention, FIG. 2 is a diagram showing waveforms of output signals and sampling timings, and FIGS. 3 and 4 are feedthrough parts of output signals. It is a figure which shows a mode that the short period noise (random noise and spike noise) was mixed in.

The image data acquisition apparatus 100 according to the embodiment of the present invention samples the CCD element 10 and the output signal of the CCD element 10 a plurality of times by the feed-through section.
A / D conversion including a high-speed A / D converter 21 that performs A / D conversion and outputs a reference signal, and a low-speed A / D converter 22 that samples once in the data section and A / D converts and outputs a data signal Means 20 and CDS processing means 30 for acquiring image data by performing correlated double sampling processing based on the reference signal and the data signal.

The CCD element 10 has a large number of light receiving pixels 11 arranged vertically and horizontally. One cycle T of the output signal S output as a voltage from the CCD element 10 is
The reset section TR indicates the voltage of the output signal when the CCD 10 is reset, and the feed through section TF is the output signal when there is no signal after the reset is released. , And the data portion TD shows the voltage corresponding to the amount of received light received by each light receiving pixel.

The high-speed A / D converter 21 samples the same feed-through section TF of the output signal S imaged by the CCD element 10 a plurality of times, for example, three times as shown in FIG. The high-speed A / D converter 21 outputs the three reference signal values R1, R2, and R3 after being D-converted. On the other hand, the low-speed A / D converter 22 samples the data portion TD once and performs A / D conversion, and the value D1 of one data signal is output from the low-speed A / D converter 22. The values R1, R2, R3 of the output reference signal and the value D1 of the data signal are CD
The image data is input to the S processing means 30 and subjected to correlated double sampling processing by the CDS processing means 30 to obtain image data.

Here, the preprocessing of the correlated double sampling processing executed by the CDS processing means 30, that is, the processing of reducing the influence of short cycle noise will be described in detail.

For example, as shown in FIG. 3, when the random noise W is mixed in the feedthrough portion, the value of the short-period noise component included in the acquired reference signal values R1, R2, and R3 is also the value. Random values are obtained, and the reference signal values R1, R2, and R3 are added and averaged to obtain one reference signal value MR (MR = (R1 + R2
When + R3) / 3) is obtained, this value MR is a value obtained by averaging the components of the random noise W included in the feedthrough portion, and is a value in which the influence of the short cycle noise component is reduced.

Further, for example, as shown in FIG. 4, when spike noise P is generated suddenly in addition to random noise W in the feedthrough portion and the spike noise P is sampled, the spike noise P is generated. The value R1 of the sampled and acquired reference signal is a unique value compared to the values R2 and R3 of the other reference signals,
The CDS processing means 30 itself detects that the value R1 of the reference signal is a peculiar value, for example, by classifying it by a threshold value, and R2 and R3 excluding R1 are averaged to obtain the value MR (MR = MR) of the reference signal. (R2 + R3) /
2) is required. The value MR is a value in which the component of the spike noise P included in the feed-through portion is removed and the component of the random noise W is averaged to reduce the influence of short cycle noise.

The reference signal obtained as described above is used in the post-processing of the correlated double sampling processing together with the data signal, and the image data obtained by the post-processing of the correlated double sampling processing is the CDS processing means 30.
Is output from the display device 50 and converted into a video signal and displayed on the display unit 50.

In the second embodiment, as shown in FIG.
The / D conversion means is a single high-speed A / D converter 23 different from that of the first embodiment, and this single high-speed A / D converter 2 is used.
Reference numeral 3 denotes a configuration in which the output signal of the CCD element 10 is sampled a plurality of times by the same feed-through section and the same data section, and A / D converted. The other configurations of the second embodiment are the same as those of the first embodiment, and the same reference numerals will be used for the same configurations as the first embodiment and the description thereof will be omitted.

The high-speed A / D converter 23 samples the same feed-through portion TD of the output signal output from the CCD element 10 a plurality of times, for example, three times, and performs A / D conversion to generate three reference signals R1 and R2. While R3 is output, the same high-speed A / D converter 23 samples the same data part TD a plurality of times, for example, three times, and performs A / D conversion to generate three data signals D1, D2, D3.
Is output (see FIG. 6). The three reference signals R1, R2, R3 and the three data signals D output above
1, D2, D3 are input to the CDS processing means 30,
By the CDS processing means 30, the preprocessing of the correlated double sampling processing for averaging the random noise and removing the spike noise component is performed on the data signal as well as the reference signal as in the first embodiment. After that, post-processing of correlated double sampling processing is performed in the same manner as above, and image data is acquired.

The image data obtained as described above is output from the CDS processing means 30, converted into a video signal and displayed on the display 50.

The plurality of times of sampling is not limited to three times, but may be any number of times as long as it is two times or more. Also,
The number of times of sampling may be different between the feedthrough section and the data section.

Further, the preprocessing of the correlated double sampling processing is not limited to the processing of removing the averaging and the peculiar value, and may be any processing so long as the short cycle noise is reduced.

[Brief description of drawings]

FIG. 1 is a block diagram showing a schematic configuration of an image data acquisition device according to an embodiment of the present invention.

FIG. 2 is a diagram showing waveforms of output signals and sampling timings.

FIG. 3 is a diagram showing a state in which short-period random noise is mixed in the feedthrough portion.

FIG. 4 is a diagram showing a situation in which short-period spike noise is mixed in the feedthrough portion.

FIG. 5 is a block diagram showing a schematic configuration of an image data acquisition device according to a second embodiment.

FIG. 6 is a diagram showing waveforms of output signals and sampling timings.

[Explanation of symbols]

10 CCD element 20 A / D conversion means 21 High-speed A / D converter 22 Low speed A / D converter 30 CDS processing means 100 image data acquisition device

   ─────────────────────────────────────────────────── ─── Continued front page    F term (reference) 5B047 AA11 AB04 BB02 CB05 DA06                       DB01 DC11                 5C024 AX01 CX06 GY01 HX23                 5C077 LL02 MM03 MP01 PP10 PP47                       PP80 PQ20 RR01 TT10                 5J022 AA01 BA02 CA10 CE01 CF10

Claims (4)

[Claims] 1. An output signal obtained by imaging with a CCD device is sampled by a feedthrough section and A / D converted to obtain a reference signal, and a data section is sampled and A / D converted to obtain a data signal. Then, in the image data acquisition method for acquiring the image data by performing the correlated double sampling processing based on the reference signal and the data signal, the reference signal is sampled a plurality of times in the same feed-through section and A / D converted. A method of acquiring image data, characterized in that 2. A CCD element and an output signal of the CCD element are sampled by a feed through section and A / D converted to output a reference signal, and a data section is sampled and A / D converted to output a data signal. In the image data acquisition device, the image data acquisition device includes: an A / D conversion unit for performing the correlated double sampling process based on the reference signal and the data signal to acquire image data. , The output signal of the CCD element,
Sampling multiple times with the same feed-through section
An image data acquisition device, which is configured to perform D / D conversion and output the reference signal. 3. A first A / D for outputting the reference signal by the A / D conversion means sampling the output signal of the CCD element a plurality of times by the same feed-through section and performing A / D conversion. Conversion means, and second A / D conversion means for performing A / D conversion at a lower speed than the first A / D conversion means for sampling once in the data section, A / D converting and outputting a data signal The image data acquisition device according to claim 2, wherein the image data acquisition device comprises: 4. The A / D conversion means samples the output signal of the CCD element a plurality of times by the same feed-through section and the same data section, respectively, and performs A / D conversion.
The image data acquisition device according to claim 2, wherein the image data acquisition device performs D conversion.