JPH01256817A - A/d converter - Google Patents

Abstract

PURPOSE:To obtain a uniform quantization accuracy over the entire signal level range of a video signal by providing a weighting means to an A/D converter of a video signal in an image pickup device or the like, and applying different weighting to each quantization step of A/D conversion. CONSTITUTION:A ladder resistance circuit 1 in an A/D converter such as N-bit full parallel comparison system consists of 2N sets of resistors. Moreover, the potential Vref across the ladder resistor is controlled variably by using a control signal of a gain control circuit GCC. Then ladder resistances R1, R2...R2N are determined to supply an equal quantization step DELTAV after nonlinear processing such as gamma characteristic or white level compression characteristic. Digital outputs D0, D1...DN-1 from a decoder circuit 3 obtained in such a way are subject to nonlinear processing, then no new gamma correction circuit or white level compression circuit is required. Thus, the resistance values of the ladder resistor circuit of the A/D conversion are made correspondent to the nonlinear characteristic in such a way, then almost uniform quantization accuracy over the entire signal level range of the video signal is obtained.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an A/D converter for video signals in an imaging device or the like.

[Conventional technology]

There are many known examples of conventional A/D converters for video signals, and recently, as the signal band has become wider, the A/D conversion speed has been improved to several tens of MHz, and 8-bit quantization has become practical. has been made into

[Problems to be Solved by the Invention] However, in A/D conversion of a video signal in an imaging device, the conventional number of quantization bits is insufficient for the following reasons.

(1) The dynamic range of the video signal is many times larger than that of a normal television signal.

(2) When gamma correction is applied after A/D conversion, quantization noise in low-luminance areas becomes three times or more.

(3) When performing white balance after A/D conversion, a dynamic range that is twice as large is required.

(4) When applying AGC after A/D conversion, a dynamic range several times larger is required.

As mentioned above, due to the insufficient number of quantization bits, conventional analog signal processing such as white compression, gamma correction, and white balance is performed before A/D conversion (this only increases the circuit scale). Moreover, the effect of digital signal processing was also halved.

Naturally, increasing the number of n-bits requires a huge increase in circuit scale, and the A/D conversion speed is also insufficient, which is extremely difficult with current A/D conversion technology and LSI technology. .

[Means for solving problems]

The A/D converter of the present invention has weighting means for weighting each quantization step of A/'D conversion differently in the A/D converter for a video signal.

[Effect]

The present invention provides uniform quantization accuracy over the entire signal level range of a video signal by weighting each quantization step of A/D conversion differently.

〔Example〕

FIG. 1 shows the first embodiment of the present invention, and shows an example of an N-bit all-parallel comparison A/D converter, where 1 is a ladder resistor circuit with 2N resistors. Consisted of. Further, the potential Vref across the ladder resistor is variably controlled by a control signal from a gain control circuit GCC. Reference numeral 2 denotes a comparison circuit, which is composed of (2N-1) comparators. 3 is a decoder circuit that converts the output of the comparison circuit 2 into an N-bit natural binary code.

Figure 2 shows the relationship between the ratio (relative value) of the resistance values of the 2N resistors of the ladder resistance circuit 1 in Figure 1 and the quantization step of the signal, where the horizontal axis represents the ladder resistance value and the vertical axis represents the quantization step. In addition, Fig. 2 also shows the non-linear characteristics of the video signal, where part A in Fig. 2 shows the characteristic corresponding to the gamma characteristic of the video signal, and part B in Fig. 2 shows the characteristic corresponding to the white compression characteristic. .

In this embodiment, as is clear from FIG. 2, the ladder resistance values R1, R2, R3 are
..., R2・-1, R2・ are determined. Of course, these resistance values are within the range where the desired nonlinear characteristics can be obtained.
For example, there is no problem in using three to four types of resistance values.

Digital output D from the decoder circuit 3 obtained in this way
O, DI, . . . DN-1 undergoes non-linear processing, so no new gamma correction circuit or white compression circuit is required.

As explained above, in this embodiment, by making the resistance value of the A/D conversion ladder resistance circuit correspond to non-linear characteristics, it is possible to obtain almost uniform quantization accuracy over the entire signal level range of the video signal. .

[Other Examples]

FIG. 3 shows a second embodiment of the present invention, in which reference numeral 11 denotes a ladder resistance circuit, which is composed of 2N resistors. There are two types of resistance values, similar to general A/D converters using the all-parallel comparison method, and as shown in Figure 3, it is composed of (2N-2) resistors R and two - resistors. There is. Further, the potential Vref across the ladder resistor is controlled by a gain control circuit GCC. I2 is a comparison circuit (2N-1,
) comparators. A decoder circuit 13 converts the output of the comparison circuit I2 into an N-bit natural binary code. 14°15.16 is a constant current circuit.

Figure 4 shows the reference voltage V generated in the ladder resistance circuit of Figure 3.
+ , V 2 , V 3 ・= V; u−k・
-V 21-Z -V2'-m, -V 21-+
.

It shows the relationship between Vref and quantization step.

At the same time, it shows the non-linear characteristics of the video signal, where part A shows gamma characteristics approximating a trifold line, and part B shows white compression characteristics.

In this embodiment, as is clear from FIG. 4, after nonlinear processing such as gamma characteristics and white compression characteristics, the reference voltage value is vl + V2, V3...V
2r- to -V 21- t -V 2'-m , -V
2I-2, V 2--+ are defined. These reference voltage values can be arbitrarily set by the current values of the constant current sources 14, 15, and 1.6. For example, if the current value II of the constant current source 14 is determined so that the current value between terminals b and c in FIG. . Similarly, between terminals c and c1,
The current value 12 of the constant current source 15.
．． All you have to do is decide r3.

As explained above, in this embodiment, a constant current source is provided at a predetermined terminal of the A/D conversion ladder resistance circuit, and the current value of this constant current source is set so as to obtain the desired nonlinear characteristics. By doing so, substantially uniform quantization accuracy can be obtained over the entire signal level range of the video signal. Furthermore, in this embodiment, if a constant current source is provided externally, it is possible to arbitrarily change the nonlinear characteristics.

Further, the constant current source may be replaced with a resistor without any problem.

Further, in the embodiment, an all-parallel comparison type A/D converter has been described, but the present invention is not limited to any type as long as it uses a ladder resistance circuit.

Furthermore, in this embodiment, a one-channel A/D converter has been described, but a plurality of color signal processing channels φ of an image pickup signal φ are used.
The A/D converter of the present invention is provided in each of the channels, and the gain control circuit GCC provided for each color signal is relatively controlled by a control signal for white balance adjustment, so that the A/D converter of each color signal channel φ is adjusted. Vref of the D converter may be relatively variably controlled.

〔Effect of the invention〕

As explained above, by weighting the quantization step of the A/D converter differently depending on the signal level,
Uniform quantization accuracy can be obtained over the entire signal level range of the video signal, and quantization noise does not increase in low signal level parts.

Further, in the present invention, desired nonlinear characteristics such as gamma characteristics and white compression characteristics can be obtained at the same time, so that a large amount of circuits can be reduced.

When the present invention is applied to a color imaging device, the white balance adjustment of color signals, for example, R, G, and B signals can be performed in advance and then inputted to the A/D converter of the present invention. be able to.

[Brief explanation of the drawing]

Figure 1 shows the first embodiment of the present invention, and is a circuit diagram of an A/D converter. Figure 2 is a diagram showing the relationship between the resistance value of the ladder resistor in Figure 1 and the quantization step. Figure 3 4 is a circuit diagram of an A/D converter according to a second embodiment of the present invention, and FIG. 4 is a diagram showing the relationship between the reference voltage and quantization step in FIG. 3. 1.11・・・・・・・・・・・・・・・・・・
・・・・・・・・・・・・・・・Ladder resistance circuit, 2
．． 12・・・・・・・・・・・・・・・・・・・・・
・・・・・・・・・・・・Comparison circuit, 3.13・
・・・・・・・・・・・・・・・・・・・・・・・・
・・・・・・・・・・・・Decoder circuit, 14.15.1
6・・・・・・・・・・・・・・・・・・・・・・・・
...Constant current source.

Claims (3)

[Claims] (1) An A/D converter for video signals, characterized by having weighting means for weighting each quantization step of A/D conversion differently. (2) The different weightings are based on the gamma characteristics of the video signal.
The A/D converter described in ). (3) The different weightings are based on the white compression characteristics of the video signal.
) The A/D converter described in item 1.