JPH0541797A - Image signal processor - Google Patents

Abstract

(57) [Abstract] [Purpose] Quantization can be performed accurately even if there is variation in the photoelectric conversion unit, and the final output with the required resolution can be easily and rapidly performed without increasing the circuit scale. It is possible to obtain. [Structure] The gain and / or offset of the amplification unit 2 is controlled so that the level range of the analog electric signal from the photoelectric conversion unit 1 matches the input level range of the A / D converter 3 at the output level of the amplification unit 2. I am trying. As a result, even if the level range of the analog electric signal from the photoelectric conversion unit 1 is narrower than the input level range of the A / D converter 3, the amplification unit 2 sets the input level range of the A / D converter 3 within the range. Even if the resolution can be matched and the resolution of the A / D converter 3 is about the same as the resolution required for the final output, it is not necessary to reduce the resolution.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image signal processing apparatus for converting an image signal into an analog electric signal by a photoelectric conversion unit, performing quantization processing on the analog electric signal and outputting the analog electric signal.

[0002]

2. Description of the Related Art Conventionally, an image signal processing device as shown in FIG. 4 has been known. In this image signal processing device, a plurality of photoelectric conversion elements are arranged for one line, and a photoelectric conversion unit 71 (linear sensor) that outputs a voltage proportional to the intensity of input light as an analog electric signal, and a photoelectric conversion unit 71. A buffer amplifier 72 for impedance-converting an analog electric signal from the A / D converter, an A / D converter 73 for quantizing the output of the buffer amplifier 72, and an operation for performing a numerical operation on the quantized output from the A / D converter 73. The processing unit 74, the black line memory 75 that holds the black correction data necessary for the black correction in the arithmetic processing unit 74, and the white correction data necessary for the white correction in the arithmetic processing unit 74 are held. White line memory 76.

Next, the operation of the image signal processing apparatus having such a configuration will be described. In this image signal processing device, black correction data and white correction data are written in the black line memory 75 and the white line memory 76, respectively, prior to the actual reading of the image signal. Therefore, first of all,
A black image signal is input to the photoelectric conversion unit 71 as a linear sensor. The black image signal input to the photoelectric conversion unit 71 is converted into an analog electric signal by the photoelectric conversion unit 71, then impedance conversion is performed by the buffer amplifier 72, and then converted into quantized data by the A / D converter 73. .. The arithmetic processing unit 74 writes the quantized data for the black image signal for one line as black correction data in the black line memory 75 as it is. This allows the black line memory 75 to hold the black correction data for one line.

Next, a white image signal is input to the photoelectric conversion unit 71. The white image signal input to the photoelectric conversion unit 71 is converted into an analog electric signal by the photoelectric conversion unit 71, then impedance conversion is performed by the buffer amplifier 72, and then converted into quantized data by the A / D converter 73. .. The arithmetic processing unit 74 once writes the quantized data for the white image signal for one line into the white line memory 76. After that, the arithmetic processing unit 74 sequentially reads the quantized data for one line for the white image signal from the white line memory 76, and the maximum value required at the final output 79 is the quantized data that is sequentially read. The quotient for one line of the calculated result is obtained as white correction data, and this is written again in the white line memory 76.

As described above, the black line memory 75,
After writing the black correction data and the white correction data in the white line memory 76, the actual image signal is read and this is processed as follows. That is, when an image signal is input to the photoelectric conversion unit 71, this image signal is input to the photoelectric conversion unit 7
1 is converted into an analog electric signal, and the buffer amplifier 7
The impedance is converted by 2 and then converted into quantized data by the A / D converter 73. In the arithmetic processing unit 74, when the image signal for one line is read, the black correction data corresponding to each position in the line of the read quantized data for one line is read from the black line memory 75, and the read quantization is read. Black correction data is subtracted from the data to perform black correction. Next, the white correction data is read from the white line memory 76, and each line of the quantized data after black correction is multiplied by the corresponding white correction data, and this is output as the final output 79. ..

By performing such a correction in the arithmetic processing unit 74, even if there is a variation in the characteristics among the photoelectric conversion elements for one line of the photoelectric conversion unit 71, the influence due to this is corrected and the image signal is corrected. Final output of the reading result of 79
Can be output as

[0007]

By the way, in the image signal processing apparatus having the above-mentioned configuration, in order to surely process the input image signal, the input voltage range of the A / D converter 73 is set to photoelectric conversion. It must be wider than the output voltage range of the section 71. Moreover, since the photoelectric conversion unit 71 is usually composed of a plurality of photoelectric conversion elements, it is necessary to further widen the input voltage range of the A / D converter 73 when the variation in the characteristics of each element is considered.

For example, when the output voltage range of the photoelectric conversion unit 71 is 0.3 V to 1.3 V and the variation of the characteristics of each element is ± 0.2 V, the maximum output voltage range is 0.
Since it becomes 1V to 1.5V, if the input voltage range of the A / D converter 73 is made to coincide with this and is set to 0.1V to 1.5V,
The maximum output voltage range of the photoelectric conversion unit 71 can be covered, and even for the output voltage of the maximum output voltage range 0.1V to 1.5V, this is accurately quantized with, for example, 8-bit resolution. be able to.

However, when the input voltage range of the A / D converter 73 is set to match the maximum output voltage range of the photoelectric conversion unit 71, the output voltage from the photoelectric conversion unit 71 is the minimum output voltage. When the range is 0.5 V to 1.1 V, the final output 79 can obtain only the resolution of 3/7 (= 0.6 V / 1.4 V) of the resolution of the A / D converter 73. Therefore, the required resolution (eg 8 bits)
In order to secure the above, it is necessary to use the 10-bit A / D converter 73 in the past.

In other words, the greater the variation in the characteristics between the photoelectric conversion elements of the photoelectric conversion unit 71, the higher the resolution of the A / D converter 73 required to secure the required resolution. Accordingly, each bit is required also in the arithmetic processing unit 74, the black line memory 75, and the white line memory 76. As a result, in an image signal processing system that requires high resolution, it is difficult to increase the speed, and there is a problem that the overall circuit scale becomes large.

According to the present invention, it is possible to accurately quantize even if there is a variation in the photoelectric conversion section, and to easily and quickly obtain the final output with the required resolution without increasing the circuit scale or the like. It is an object of the present invention to provide an image signal processing device capable of

[0012]

In order to achieve the above object, an image signal processing device of the present invention comprises a photoelectric conversion means for converting an image signal into an analog electric signal, and an analog electric signal outputted from the photoelectric conversion means. Amplifying means for amplifying, a quantizing means for inputting an analog electric signal as a result of being amplified by the amplifying means and quantizing the analog electric signal into quantized data, and a gain and / or offset of the amplifying means are made variable. Control means for controlling the control means, wherein the control means controls the level of the analog electric signal from the photoelectric conversion means to match the input level range of the quantizing means at the output level of the amplifying means. It is characterized in that it is adapted to control the gain and / or the offset.

Further, the amplifying means is characterized by having a function as a differential amplifier for amplifying a difference between the analog electric signal output from the photoelectric converting means and a constant bias.

Further, the amplifying means has a function as a differential amplifier for amplifying a difference between the analog electric signal output from the photoelectric converting means and the dummy signal superimposed on the analog electric signal. Is characterized by.

[0015]

According to the present invention, the gain and / or offset of the amplifying means is controlled so that the level range of the analog electric signal from the photoelectric converting means matches the input level range of the quantizing means at the output level of the amplifying means. ing. As a result, even if the level range of the analog electric signal from the photoelectric conversion means is narrower than the input level range of the quantizing means, the amplifying means can match the input level range of the quantizing means. Even if the resolution of the digitizing means is similar to that required for the final output,
It is not necessary to reduce the resolution.

When the amplifying means has a function as a differential amplifier for amplifying the difference between the analog electric signal output from the photoelectric converting means and a constant bias, the amplifying means has a small input dynamic range. Can be
The saturation of the amplifier can be prevented.

Further, when the amplifying means has a function as a differential amplifier for amplifying the difference between the analog electric signal output from the photoelectric converting means and the dummy signal superimposed on the analog electric signal, , Fixed pattern noise and the like included in the dummy signal can be removed from the analog electric signal.

[0018]

Embodiments of the present invention will be described below with reference to the drawings. 1 is a block diagram of a first embodiment of an image signal processing apparatus according to the present invention. The image signal processing device shown in FIG. 1 includes a photoelectric conversion unit 1 that outputs a voltage proportional to the intensity of input light as an analog electric signal, an amplification unit 2 that amplifies the output voltage from the photoelectric conversion unit 1, and an amplification unit. An A / D converter 3 for quantizing the output from the unit 2, an arithmetic processing unit 4 for performing a numerical operation on the quantized output from the A / D converter 3,
The black line memory 5 that holds the black correction data necessary for the black correction in the arithmetic processing unit 4 and the white line memory 6 that holds the white correction data necessary for the white correction in the arithmetic processing unit 4 And the latches 7 and 8, and the gain control means 10 for controlling the gain G of the amplification section 2 is provided.
And an offset control means 11 for controlling the offset voltage V _OFS are provided, the amplification unit 2 is an amplifier in which the gain G and the offset voltage V _OFS are variable, and the latches 7 and 8 are provided with arithmetic processing. The gain G and the offset voltage V _OFS determined by the unit 4 are set.

In the first embodiment, under the configuration as described above, the output voltage range of the photoelectric conversion unit 1 is controlled by controlling the gain G of the amplification unit 2 and the offset voltage V _OFS . The output level is made to match the input voltage range of the A / D converter 3, and in this state, the black correction data and the white correction data are taken in, and then the black correction data and the white correction data are used for the actual image signal. It is designed to perform arithmetic processing.

Next, the operation of the image signal processing apparatus having such a configuration will be described more specifically. In the following,
It is assumed that the photoelectric conversion unit 1 is composed of a plurality of photoelectric conversion elements for one line, the range of the output voltage of the photoelectric conversion unit 1 is 0.3 V to 1.3 V, and the variation of the characteristics of each photoelectric conversion element. Is ± 0.2 V. Also, the final output 9
8 bit resolution is required, and A / D converter 3
The input voltage range of this A / D converter 3 is 0 V (= V _AD , _MIN ) to 1.6 V.
(= V _AD , _MAX ). That is, in such an A / D converter 3, a minimum voltage 0V (= V _AD ,
_{When MIN} ) is input, "0" is output as quantized data, and when the maximum voltage of 1.6 V (= V _AD , _MAX ) is input, "255" is output as quantized data. There is.

Now, assuming that the output voltage of the photoelectric conversion unit 71, that is, the input voltage of the amplification unit 2 is V _IN , and the gain and offset voltages are G and V _OFS , respectively, the output voltage V _OUT from the amplification unit 2 is Given by the formula.

[0022]

[Equation 1]

Therefore, first, the output voltage V of the photoelectric conversion unit 1
_The gain G is adjusted so that the range of the output voltage V _OUT as a result of the amplification of _IN by the amplification unit 2 falls within the input voltage range of the A / D converter 3.
And the offset voltage V _OFS is initialized. In this example, the maximum output voltage range of the photoelectric conversion unit 1 is 0.1V to 1.V.
Since the voltage is 5 V, initially, in the amplifying unit 2, if the gain G is initialized to 1 time and the offset voltage V _OFS is initialized to 0 V, all the output voltage V _OUT of the amplifying unit 2 is A / D.
It falls within the input voltage range of the converter 3. Therefore, the arithmetic processing unit 4 initializes the initial value G _INT of the gain G and the initial value D _INT of the offset voltage V _{OFS in} the latches 7 and 8, respectively. The initial value G _INT of the gain G and the initial value D of the offset voltage V _OFS set in the latches 7 and 8 respectively.
_INT is read by the gain control unit 10 and the offset control unit 11, respectively, and is used for gain control and offset voltage control of the amplification unit 2 based on the equation (1).

After initializing the gain G and the offset voltage V _OFS as described above, first, the black image signal is taken in by the photoelectric conversion unit 1, amplified by the amplification unit 2, and then by the A / D converter 3. Quantize and output to the arithmetic processing unit 4. The arithmetic processing unit 4 writes this quantized data in the black line memory 5, and when this process is repeated for one line, the minimum value D is obtained from the quantized data for one line held in the black line memory 5. _{Find MIN} . In addition, D represents digital data, that is, quantized data as a result of analog-digital conversion of the analog voltage signal V by the A / D converter 3.

Next, the white image signal is taken in by the photoelectric conversion unit 1, amplified by the amplification unit 2, quantized by the A / D converter 3 and output to the arithmetic processing unit 4. In the arithmetic processing unit 4,
This quantized data is written in the white line memory 6, and when this process is repeated for one line, the maximum value D _MAX is obtained from the quantized data for one line held in the white line memory 6.

In this way, when the minimum value D _MIN of the quantized data held in the black line memory 5 and the maximum value D _MAX of the quantized data held in the white line memory 6 are obtained, these are respectively obtained. , The lower limit value and the upper limit value of the output voltage range of the photoelectric conversion unit 1 are the output voltage range of the amplification unit 2 as a result of amplification in the amplification unit 2 with the initial value G _INT of the gain and the initial value D _INT of the offset voltage. Therefore, the arithmetic processing unit 4 calculates and amplifies the gain G and the offset voltage V _OFS of the amplification unit 2 so that the output voltage range of the amplification unit 2 matches the input voltage range of the A / D converter 3. A new output voltage range (V _AD , _MIN ) that matches the current output voltage range (D _{MIN to} D _MAX ) of the unit 2 with the input voltage range (V _AD , _{MIN to} V _AD , _MAX ) of the A / D converter 3. To V _AD , _MAX ).
That is, when the lower limit value of the output voltage range of the photoelectric conversion unit 1 is input to the amplification unit 2 as the input voltage V _IN (= D _IN ),
The output voltage of the amplifier 2 is from D _MIN to V _AD , _MIN (= “0”)
In addition, when the upper limit value of the output voltage range of the photoelectric conversion unit 1 is input to the amplification unit 2 as the input voltage V _IN (= D _IN ), the output voltage of the amplification unit 2 changes from D _MAX to V _AD , _MAX (=
"255"), the gain G and the offset voltage V _OFS are calculated according to the following equation 1 according to the following equations.

[0027]

[Equation 2]

Then, these are transferred to the latches 7 and 8, respectively, and replaced with the initial value G _{INT of the} gain and the initial value D _INT of the offset voltage held in the latches 7 and 8 and held in the latches 7 and 8. When the new gain G and offset voltage V _OFS are set in the latches 7 and 8 as described above, the gain control unit 10 and the offset control unit 11 cause the gain G and offset voltage V newly set from the latches 7 and 8, respectively. _{The OFS} is read and the gain and offset voltage of the amplifier 2 are controlled based on the equation 1.

The following table shows the results when the above control is performed for each of the minimum, standard, and maximum output voltage ranges of the photoelectric conversion unit 1 which may occur due to variations in the characteristics of the photoelectric conversion elements. In addition, in the following table, V _OFS , G
, Are rounded down to the second decimal place and the third decimal place.

[0030]

[Table 1]

As is clear from the results of Table 1, the data D _MIN , D _MAX obtained by taking in the black image signal and the white image signal.
Based on _Equation 2, gain G, offset voltage V _OFS
By controlling the amplification unit 2 by this, the output voltage range of the photoelectric conversion unit 1 can be matched with the input voltage range of the A / D converter 3 at the output level of the amplification unit 2.

After setting the gain G and the offset voltage V _OFS in the latches 7 and 8 in this manner, the photoelectric conversion unit 1
, The black image signal is taken in again, the output voltage of the photoelectric conversion unit 1 is amplified by the gain G and the offset voltage V _OFS newly set in the latches 7 and 8, and quantized by the A / D converter 3 to obtain the quantized data. Is written in the black line memory 5 as black correction data, and this is repeated for one line.

Next, the white image signal is fetched again from the photoelectric conversion unit 1, the quantized data is once written in the white line memory 6 by the same procedure as described above, and this is repeated for one line. Then, the arithmetic processing unit 4 sequentially reads the quantized data for one line for the white image signal from the white line memory 6 and outputs the maximum value "25" required for the final output 9.
The quotient for one line, which is the result of calculating 5 ″ by the above-mentioned sequentially read quantized data, is obtained as white correction data, and this is written in the white line memory 6 again.

As described above, after writing the black correction data and the white correction data in the black line memory 5 and the white line memory 6, respectively, the actual image signal is read and processed as follows. That is, when an image signal is input to the photoelectric conversion unit 1, this image signal is converted into an analog electric signal, that is, a voltage in the photoelectric conversion unit 1. The output voltage of the photoelectric conversion unit 1 is amplified by the amplification unit 2 with the gain G and the offset voltage V _OFS held in the latches 7 and 8, and then converted into quantized data by the A / D converter 73. .. When the arithmetic processing unit 4 reads an image signal for one line, the black correction data corresponding to each position in the line of the read quantized data for one line is read from the black line memory 5, and the read quantization is read. Black correction data is subtracted from the data to perform black correction. Next, the white correction data is read from the white line memory 6, each line of quantized data after black correction is multiplied by the corresponding white correction data, and this is output as the final output 9. ..

By such a correction being performed in the arithmetic processing unit 74, even if there is a variation in characteristics among the photoelectric conversion elements for one line of the photoelectric conversion unit 1, the influence due to this is corrected and the image signal is corrected. The read result of can be output as the final output 9.

By the way, in this embodiment, the photoelectric conversion unit 1
The maximum output voltage range is 0.1V to 1.5
Regardless of whether the output voltage is V or the minimum output voltage range is 0.5V to 1.1V, the gain G and the offset voltage V _OFS of the amplification unit 2 are appropriately controlled so that the amplification unit Since the output voltage of 2 can always match the input voltage range of the A / D converter 3, the A (A) having a resolution (8 bits) similar to that required for the final output 9 (8 bits). The / D converter 3 can be used to obtain the final output 9 with the required resolution (8 bits). In other words, it is not necessary to set the resolution of the A / D converter 3 to a high resolution regardless of how much the characteristics of the photoelectric conversion elements of the photoelectric conversion unit 1 vary. With the A / D converter 3, it is possible to always secure a resolution of 8 bits as the resolution of the final output 9.

Further, as a result, it is not necessary to increase the number of bits in the arithmetic processing unit 4, the black line memory 5 and the white line memory 6, and as a result, when constructing an image processing system that requires high resolution. Also, it is easy to increase the processing speed, and the overall circuit scale can be kept small.

FIG. 2 is a block diagram of the second embodiment of the image signal processing apparatus according to the present invention. In FIG. 2, the same parts as those in FIG. 1 are designated by the same reference numerals. In the second embodiment, a bias unit 23 that generates a constant bias voltage V _BS is further provided, and a differential amplifier is used as the amplification unit 22. That is, the amplifier 22 outputs the output voltage V from the photoelectric conversion unit 1 and the constant bias voltage V from the bias unit 23.
_BS and _BS are input, the difference (V- _VBS ) is differentially amplified and given to the A / D converter 3.

Even in such a configuration, the operation similar to that in the first embodiment can be performed and the effect obtained in the first embodiment can be obtained. However, in the second embodiment, the amplification is further performed. Since the section 22 is adapted to perform differential amplification between the output voltage V of the photoelectric conversion section 1 and the constant bias voltage V _BS , the input dynamic range required in the amplification section 22 is the same as in the first embodiment. The input dynamic range required for the amplification unit 2 is smaller than that required, and the saturation of the amplifier can be effectively prevented.

FIG. 3 is a block diagram of a third embodiment of the image signal processing apparatus according to the present invention. In the third embodiment, the same differential amplifier as that of the second embodiment is used for the amplifier 32, but the amplifier 32 has the output voltage V from the photoelectric conversion unit 1 and the output voltage V. Superimposed dummy signal V _DMY
Are inputted, the difference (V-V _DMY ) is differentially amplified and given to the A / D converter 3.

With such a configuration, the dummy signal component superimposed on the output voltage V of the photoelectric conversion section 1 can be effectively removed. For example, when the photoelectric conversion unit 1 is a linear sensor such as a CCD, by removing the dummy signal component, it is possible to remove the fixed pattern noise and the like contained therein.

In each of the above-mentioned embodiments, the amplification units 2, 2
After the gain G of 2 or 32 and the offset voltage V _OFS are determined, the gain G and the offset voltage V _OFS are set to be constant, but when the capture is performed, the black correction data and the white correction data corresponding to the photoelectric conversion element are acquired. It is also possible to sequentially read and set the gain G and the offset voltage V _OFS each time to perform signal processing. In this case, the image signal processing can be performed with higher accuracy.

[0043]

As described above, according to the present invention,
The gain and / or offset of the amplifying means is controlled so that the level range of the analog electric signal from the photoelectric converting means matches the input level range of the quantizing means at the output level of the amplifying means. Even if there are variations in the values, the quantization can be performed accurately, and the final output with the required resolution can be obtained easily and at high speed without increasing the circuit scale or the like.

Further, when the amplifying means has a function as a differential amplifier for amplifying the difference between the analog electric signal output from the photoelectric converting means and a constant bias, the amplifying means has a small input dynamic range. Can be
The saturation of the amplifier can be prevented.

Further, when the amplifying means has a function as a differential amplifier for amplifying a difference between the analog electric signal output from the photoelectric converting means and the dummy signal superimposed on the analog electric signal, , Fixed pattern noise and the like included in the dummy signal can be removed from the analog electric signal.

[Brief description of drawings]

FIG. 1 is a block diagram of a first embodiment of an image signal processing device according to the present invention.

FIG. 2 is a block diagram of a second embodiment of the image signal processing device according to the present invention.

FIG. 3 is a block diagram of a third embodiment of the image signal processing device according to the present invention.

FIG. 4 is a block diagram of a conventional image signal processing device.

[Explanation of symbols]

 DESCRIPTION OF SYMBOLS 1 Photoelectric conversion part 2,22,32 Amplification part 3 A / D converter 4 Arithmetic processing part 5 Black line memory 6 White line memory 7,8 Latch 10 Gain control means 11 Offset control means 23 Bias part

Claims (3)

[Claims] 1. A photoelectric conversion means for converting an image signal into an analog electric signal, an amplification means for amplifying the analog electric signal output from the photoelectric conversion means, and an analog electric signal resulting from the amplification by the amplification means are inputted. Quantizing means for quantizing the analog electric signal into quantized data, and control means for variably controlling the gain and / or offset of the amplifying means, wherein the control means is provided from the photoelectric conversion means. An image signal, characterized in that the gain and / or offset of the amplification means is controlled so that the level range of the analog electric signal matches the input level range of the quantization means at the output level of the amplification means. Processing equipment. 2. The image signal processing device, wherein the amplification means has a function as a differential amplifier for amplifying a difference between an analog electric signal output from the photoelectric conversion means and a constant bias. .. 3. The amplifying means has a function as a differential amplifier for amplifying a difference between an analog electric signal output from the photoelectric converting means and a dummy signal superimposed on the analog electric signal. An image signal processing device characterized by: