KR19980027326A - TV Display Contrast Enhancement Circuit Using Adaptive Histogram Equalization - Google Patents

Abstract

The present invention relates to a TV screen contrast enhancement circuit using an adaptive histogram equalization method for actively contrasting the contrast ratio according to an input signal of a TV to improve the contrast on a dark background and the contrast on a bright background. In the related art, the screen contrast is changed according to the change of the DC level of the luminance signal because only the amplitude is increased to improve the screen contrast.

In other words, the contrast on a light background becomes smaller, and the contrast on a dark background becomes larger. Therefore, if the contrast on the dark background is improved, the brightness becomes saturated on a bright screen, resulting in deterioration of image quality. There was a problem that the contrast of the screen on a dark background does not appear much by adjusting the contrast ratio.

The present invention improves the screen contrast on a dark background and the screen contrast on a bright background by varying the screen contrast ratio using a histogram according to an actively input signal. It is intended to watch video, and it is applied to a VCR that separates and outputs luminance and color signals.

Description

TV Display Contrast Enhancement Circuit Using Adaptive Histogram Equalization

The conventional TV's screen contrast adjustment circuit, as shown in Figure 1, the microcomputer (1) for outputting the screen contrast signal by the adjustment of the internal IIC, and receives the screen contrast signal output from the microcomputer (1) internal amplifier It is composed of an image color deflection unit 2 for amplifying the luminance signal through and outputting the image to be in a desired contrast state.

As shown in FIG. 1, the conventional TV's screen contrast adjustment circuit has an image color deflection unit when the screen contrast signal is output as shown in (a) of FIG. 2 by the adjustment of the internal IIC of the microcomputer 1. (2) receives the luminance signal, which is the screen contrast signal output from the microcomputer 1, and amplifies the screen contrast as shown in (b) of FIG. 2 through an internal amplifier, and outputs the desired screen contrast state.

Here, it can be seen that the contrast adjustment is a simple amplitude linear amplification, so that the screen contrast can be easily improved, and the screen contrast is improved in the dark areas where the amplitude of the input signal is very small.

However, in the related art, the screen contrast is changed according to the change of the DC level of the luminance signal because only the amplitude is increased to improve the screen contrast.

In other words, the contrast on a light background becomes smaller, and the contrast on a dark background becomes larger. Therefore, if the contrast on the dark background is improved, the brightness becomes saturated on a bright screen, resulting in deterioration of image quality. There was a problem that the contrast of the screen on a dark background does not appear much by adjusting the contrast ratio.

The present invention improves the screen contrast on a dark background and the screen contrast on a bright background by varying the screen contrast ratio using a histogram equalization method according to an actively input signal in order to solve such a conventional problem. When explaining the configuration and effect of the present invention by the accompanying drawings to be as follows.

1 is a block diagram of a conventional TV contrast adjustment.

Fig. 2A is a luminance signal waveform diagram input to a conventional image color deflection portion.

(B) shows the luminance signal waveform when the contrast of the conventional screen is increased.

Figure 3 is a block diagram showing the adaptive contrast enhancement circuit of the TV using the present invention adaptive histogram equalization method.

Figure 4 is a detailed block diagram showing the structure of the present invention adaptive contrast enhancement.

Figure 5 is a detailed block diagram showing the structure of the adaptive portion of the present invention adaptive contrast enhancement.

Figure 6 is a block diagram showing a detailed structure of the equalizer of the present invention.

7 is a graph showing an operating state of the band determining unit of the present invention.

Figure 8 (a) is a graph showing the equalization state for the bright image of the present invention.

(B) is a graph showing the histogram after the equalization state of the present invention.

Figure 9 (a) is a graph showing the state before equalization for the dark image of the present invention.

(B) is a graph showing the state after equalization for the dark image of the present invention.

10 is a table showing a calculation formula for explaining the present invention.

The present invention relates to a TV screen contrast enhancement circuit using an adaptive histogram equalization method for actively contrasting a contrast ratio according to an input signal of a TV to improve contrast on a dark background screen and a contrast on a bright background screen.

First, the TV screen contrast enhancement circuit using the adaptive histogram equalization method of the present invention includes an audio / video switching unit 10 for switching and outputting an input audio / video signal as shown in FIG. Equalization method of the internal histogram by receiving the image signal output from the switching unit 10 receives the luminance signal and the color signal separately and outputs, and the luminance signal filtered from the comb filter unit 11 Adaptive Contrast Enhancement Unit 12 for improving and outputting the screen contrast, and the luminance signal output from the Adaptive Contrast Enhancement unit 12 and the color signal output from the comb filter unit 11 are received by the internal amplifier. The image color deflection unit 13 outputs a screen contrast signal in a desired state.

4 is a detailed block diagram of the adaptive contrast improving unit 13 of the present invention. The clamping unit 14 for removing the synchronization of the luminance signal output from the comb filter unit 11 and the synchronization signal from the clamping unit 14 are shown. And an A / D converter 15 for converting the luminance signal outputted from the digital signal into a digital signal, a field memory unit 16 for storing each field of the luminance signal converted and outputted into the digital signal, and a field memory unit ( 16, and an adaptation unit 17 for improving the screen contrast according to the field input to the field, and a D / A converter 18 for converting the screen contrast signal output from the adaptation unit 17 into an analog signal. will be.

5 is a detailed block diagram of the adaptation unit 17, which detects the characteristics of a signal input through the field memory 16 and outputs a luminance frequency value [h (Y)] that is a histogram; A maximum value extractor 20 which receives a histogram output from the signal detector 19 and detects a luminance value YM that is most distributed in the histogram; and a luminance value output from the maximum value extractor 20 YM) and the equalization value input by the user, and a band for determining the bandwidth by comparing the luminance frequency value [h (Y)] output from the signal detector 19 with the luminance value YM as the center. The quantized probability distribution function value [h * (V *)] is input to determine the equalization rate by receiving the decision unit 21 and the bandwidth YM + T (YM-T) determined by the band determiner 21. The equalizer 22 outputs and the probability distribution function value [h * (V *)] outputted from the equalizer 22 are received and stored in the field memory 16. The signal recovery unit 23 converts the call into an equalized signal and outputs the converted signal.

FIG. 6 is a block diagram showing the detailed structure of the equalizer 22. The histogram h (Y) output from the signal detector 19 is removed to remove only the portion not to be equalized. Frequency band calculating section 24 for outputting the band cutting section 24 and the histogram [h '(Y)] output from the band cutting section 24 to calculate the probability [P (Y)] of each luminance value. 25), a cumulative distribution calculation unit 26 for calculating a cumulative distribution function V from the probability [P (Y)] output from the frequency calculating unit 25, and an output from the maximum value extracting unit 20. Determining which region among the three luminance regions receives the luminance value YM, and determines the equalization rate so as to output the luminance value suitable for the corresponding region. The unit 27 and the cumulative distribution output from the cumulative distribution calculation unit 26 by receiving the equalization rate (∝) output from the equalization rate determination unit 27. (V), such as the equalization and outputs a quantized value probability distribution functions [h * (* V)] to be composed of both section 28.

As illustrated in FIG. 3, when the video signal is switched and output from the audio / video switching unit 10, the comb filter unit 11 generates a luminance signal and a color signal of the switched video signal. Then, the luminance signal is applied to the adaptive contrast enhancement unit 12 and the color signal is applied to the image color deflection unit 13.

In this case, as shown in FIG. 4, the adaptive contrast enhancement unit 12 receives the luminance signal output from the comb filter unit 11 from the clamping unit 14, removes the synchronization signal, and outputs the A / D conversion. The unit 15 converts the luminance signal from which the synchronization signal is removed into a digital signal and outputs the digital signal.

Therefore, each field of the digital signal converted and output by the A / D conversion unit 15 is stored in the field memory unit 16 and then input to the adaptation unit 17 and stored in the field memory unit 16 and output. After improving the screen contrast according to the field, the D / A converter 18 converts the analog signal into an analog signal and outputs the converted analog signal.

Meanwhile, as shown in FIG. 5, the adaptation unit 17 receives the field signal output from the field memory unit 16 from the signal detection unit 19, grasps the characteristics of the input signal, and histogram [h (Y)]. ], The maximum value extractor 20 receives the histogram [h (Y)] and detects the luminance value YM that is most distributed in the histogram indicating the distribution of the luminance values. ) And the equalizer 22.

In this case, the band determiner 21 receives the luminance value YM output from the maximum value extractor 20 and an arbitrary equalization value input by the user, and determines and outputs a band value to be equalized among the input signals. As shown in (a) of FIG. 7, when the sum of all the values in the histogram is equal to the number of pixels N, the equalization value input by the user is received. The bandwidth (YM + T) (YM-T) is compared with the value determined by adding the histogram [h (Y)] output from the maximum value extracting unit 20 based on the luminance value YM output from (20). The output is determined.

That is, if the user decides to equalize the total number of pixels by χ%, the sum [S (T)] of the luminance frequency value [h (Y)] can be calculated to be equal to or greater than the equalization value determined by the user. As shown in FIG. 10, the sum of the frequency values [S (T)] is calculated and compared by Equation 1.

Therefore, when the sum S (T) of the frequency values satisfies the equalization value determined by the user, the value of the bandwidth YM + T (YM-T) is output to the equalizer 22.

The bandwidth may be represented as in Equation 2 shown in FIG.

As such, when the bandwidth YM + T (YM-T) determined by the band determiner 21 is output, the equalizer 22 receives the input quantized probability distribution function value [h * (V *) to determine an equalization rate. ] To the signal restorer 23, the signal restorer 23 receives the probability distribution function value [h * (V *)] output from the equalizer 22 and stores the signal stored in the field memory 16. Is converted to an equalized signal and output.

At this time, the equalizer 22 receives the histogram [h (Y)] outputted by grasping the characteristic of the input signal from the signal detector 19 and received from the band cut section 24, as shown in FIG. Only the portion that is not equalized is removed and output. The transfer function H (Y) formed by the bandwidth YM + T (YM-T) determined and output by the band determining unit 21 is shown in FIG. As shown in Equation 3, the band-limited histogram [h '(Y)] output from the band cut section 24 has a luminance value between the bandwidth (YM-T) and the bandwidth (YM + T). Will have only the value of.

On the other hand, the frequency calculator 25 receives the band-limited histogram [h '(Y)] output from the band cut section 24 and shows the probability [P (Y)] for each luminance value as shown in FIG. Calculate and output according to Equation 4.

In this way, the probability P (Y) calculated by the frequency calculating unit 25 is inputted to the cumulative distribution calculating unit 26 and calculated by Equation 5 as shown in FIG. 10 to calculate the cumulative distribution function [V or F ( Y)] is output to the back quantum unit 28.

That is, when the probability P (Y) for each luminance value is computed and output, the cumulative distribution function F (Y) is distributed between 0.1 by Equation 8 shown in FIG. See Equation 3 in the diagram)

Therefore, when the histogram value h (χi) of an arbitrary pixel is determined, the luminance value is a value between the bandwidth YM + T and the bandwidth YM-T, so that the probability P (Y) in the arbitrary pixel is determined. ] Is output by equation 9 shown in FIG.

At this time, the equalization rate determiner 27 receives the luminance value YM outputted from the maximum value extractor 20 and determines which region among the three luminance regions is the luminance value YM, which is different for each region. The three luminance areas are visually recognized even when the screen contrast is slightly changed (low luminance value is about 0-50), intermediate luminance area (luminance value is about 51-200), and the visual contrast is changed. The high luminance region (luminance value is about 201-255) insensitive to the < RTI ID = 0.0 >

In this way, the equalization rate (∝) output from the equalization rate determining unit 27 and the cumulative distribution function [V or F (Y)] output from the cumulative distribution calculating unit 26 are together at the equal quantum unit 28 for quantization. As input, the quantum unit 28 outputs the probability distribution function V * quantized by Equation 7, as shown in FIG.

Therefore, since the slope component Yband · ∝ of the probability distribution function V * output from the back quantum unit 28 is inversely proportional to the degree of equalization, the cumulative distribution output from the cumulative distribution calculation unit 26 when the slope component is large. Since the change in the probability distribution function V * output from the back quantum unit 28 is large according to the change in the function V, the equalization rate is decreased. The quantum unit 28 finally calculates the probability [Pv * and h * (V *)] according to the characteristics of the probability distribution function V * to obtain the finally calculated probability [h * (V *)]. Will print.

Here, Pv * is a sum of probabilities of luminance values having the same cumulative distribution function value, and h * (V *) is a sum of histogram values [h (v)] of luminance values having the same cumulative distribution function value.

That is, the quantum unit 28 outputs the luminance value output from the maximum value extracting unit 20 such that the probability [P (Y)] computed and output by the frequency calculating unit 25 is uniformly quantized through the cumulative distribution calculating unit 26. The luminance level of (YM) corresponds to Equation 10 of FIG. 10, which is an intermediate luminance region (more than 51 and less than 209). By increasing the probability distribution function V * and outputting the result, the equalization amount is increased, and in the case of the luminance region 210 or higher where the luminance value YM is high, equalizing the maximum 30% by Equation 12 of FIG. Will reduce the volume.

On the other hand, (a) of FIG. 8 shows the equalization amount for the bright image, and (b) shows the histogram after the equalization for the bright image.

That is, in Figure 8 (a) equalizes about 90% of the original image histogram and is equalized in a bright area, the equalization factor value is about 0.9, so that the equalization result appears as in (b) of Figure 8 The contrast of the screen in the areas where the contrast is prominent (less than the luminance value output from the maximum value extracting unit) is about twice as high.

9A is a graph showing the state before equalization of the dark image, and (B) is a graph showing the state after equalization.

That is, since the total number of pixels for dark image is 759, equalizing them to 96%, the luminance value is from 43 to 56 as shown in FIG. 9, so the bandwidth (Yband) is 13 and equalization because it is the dark image. Since the ratio is about 1.1, the result of the equalization shows that there is almost no luminance of very small value as in (b) of FIG. 9, but the distribution of the latitude is increased by equalization, which improves the screen contrast. .

In this case, the smaller the equalization rate is, the more the equalization is. When the equalization rate is 1 or more, the degree of equalization is reduced, and the contrast of the specific part (fine part) is greatly improved.

As described above, by varying the screen contrast ratio using a histogram according to an actively input signal, the reliability of the product is improved by improving the contrast of a dark background and the contrast of a bright background. It is effective.

Claims (4)

Adaptive Contrast Enhancement unit which receives the luminance signal filtered out from the comb filter unit 11 and improves the screen contrast using the equalization method of the internal histogram to output the luminance signal with improved image contrast to the image color deflection unit 13 ( 12) TV screen contrast enhancement circuit using an adaptive histogram equalization method characterized in that it is provided. The adaptive contrast enhancing unit 12 is a clamping unit 14 for removing the synchronization of the luminance signal output from the comb filter unit 11, and the synchronization signal is removed from the clamping unit 14 and output. An A / D conversion unit 15 for converting the converted luminance signal into a digital signal, a field memory unit 16 for storing each field of the luminance signal converted and outputted into the digital signal, and an input into the field memory unit 16; It is characterized in that it is composed of an adaptation unit 17 for improving the screen contrast according to the filter, and a D / A conversion unit 18 for converting the screen contrast signal output improved from the adaptation unit 17 into an analog signal TV screen contrast enhancement circuit using adaptive histogram equalization. The signal detecting unit (19) according to claim 2, wherein the adaptor (17) detects the characteristics of the signal input through the field memory (16) and outputs a luminance frequency value (h (Y)) that is a histogram. A maximum value extractor 20 that receives the histogram output from 19 and detects a luminance value YM that is most distributed in the histogram, and a luminance value YM output from the maximum value extractor 20 And a band determining unit which receives the equalization value input by the user and determines the bandwidth by adding the luminance frequency value [h (Y)] output from the signal detection unit 19 around the luminance value YM to determine the bandwidth. (21), and outputs a quantized probability distribution function value [h * (V *)] to receive the bandwidth YM + T (YM-T) determined by the band determiner 21 and determine an equalization rate. The equalizer 22 receives the probability distribution function value [h * (V *)] output from the equalizer 22 and equalizes the signal stored in the field memory 16. A TV screen contrast enhancement circuit using an adaptive histogram equalization method, characterized in that the signal recovery unit 23 converts and outputs the converted signal. 4. The equalizer 22 receives the histogram [h (Y)] output from the signal detector 19 and removes only the portions not to be equalized to output a band-limited histogram [h '(Y)]. A frequency calculating section 25 which receives a band cutting section 24 and a histogram [h '(Y)] output from the band cutting section 24 and calculates a probability [P (Y)] of occurrence of each luminance value; A cumulative distribution calculating unit 26 for calculating a cumulative distribution function V from the probability [P (Y)] output from the frequency calculating unit 25 and a luminance value YM output from the maximum value extracting unit 20. An equalization rate determiner 27 for determining which equalization rate is determined so that the luminance value YM is located in one of the three luminance areas, and determining the equalization rate so as to output a luminance value suitable for the corresponding area. And, by receiving the equalization rate (∝) output from the equalization rate determination unit 27 equalizes the cumulative distribution function (V) output from the cumulative distribution calculation unit 26 both The probability distribution function values [h * (V *)] Using the adaptive histogram equalization method, characterized in that consists of the quantization unit 28 such that the output contrast enhancement circuit TV screen.