JP3551655B2 - Gamma correction device - Google Patents

Description

[0001]
TECHNICAL FIELD OF THE INVENTION
The present invention relates to a gamma (hereinafter referred to as γ) correction device (International Patent Classification H04N 9/69) for a video signal, and is particularly effective when used for gradation correction of a printer.
[0002]
[Prior art]
Conventionally, γ correction processing is generally performed for the purpose of gradation correction of a printer. In the gamma correction processing, the input image is converted using one fixed gamma correction table regardless of the gradation characteristics of the input image data, or even in a printer that can set the gamma correction table, the operator can use the image or its image. The gamma correction table has been set while observing the gradation characteristics.
[0003]
Therefore, in order to reduce these operations, many methods for automating the γ correction processing have been proposed, for example, Japanese Patent Application Laid-Open No. Sho 60-139080. Hereinafter, the conventional gamma correction device will be described.
[0004]
FIG. 7 is a block diagram of a gamma correction circuit in a conventional example. In FIG. 7, reference numeral 601 denotes an R signal γ correction unit, 602 denotes a G signal γ correction unit, 603 denotes a B signal γ correction unit, 604 denotes a luminance signal histogram creation unit, and 605 denotes a cumulative luminance distribution curve setting unit. The operation will be described below.
[0005]
A luminance signal Y is calculated from the input signal RGB signal, and a luminance signal histogram is created by the luminance signal histogram creating means 604. The cumulative luminance distribution is obtained from the histogram, and the cumulative luminance curve is set by the cumulative luminance distribution curve setting means 605. The γ correction curve is obtained by normalizing this cumulative luminance curve to a value of 0 to 255. By using the γ correction curve for the γ correction of the R signal, the G signal, and the B signal, the γ correction according to the gradation characteristics of the image is automatically performed.
[0006]
[Problems to be solved by the invention]
However, in the above-described γ correction circuit, when the degree of backlight is large, the gradation is corrected so that the gradation becomes uniform. Key jumps become noticeable as false contours. In addition, there is a problem in that a portion having good gradation is compressed in the opposite manner and the contrast is reduced.
[0007]
[Means for Solving the Problems]
In order to solve the above-mentioned problem, a gamma correction apparatus according to the present invention includes means for creating a histogram indicating a gradation distribution of an RGB signal which is an input video signal, and detecting a gradation jump in a low gradation part from the histogram. And a means for detecting the degree of gradation loss in the low gradation part, and determining the γ correction amount from the gradation skip and gradation loss degree.
[0008]
According to the present invention, it is possible to calculate the value of γ to be corrected based on the degree of gradation loss, and to perform the optimal γ correction processing on any input image. Further, by not performing the γ correction on the image having the gradation jump, it is possible to prevent the image quality such as the false contour from deteriorating.
[0009]
BEST MODE FOR CARRYING OUT THE INVENTION
The gamma correction apparatus according to claim 1 of the present invention provides a means for creating a histogram indicating a gradation distribution of an RGB signal as an input video signal, and a means for detecting a gradation jump of a low gradation part from the histogram. Means for detecting the degree of gradation loss in the low gradation part, wherein the gamma correction is not performed when the gradation skip exists, and the gamma correction is performed based on the gradation loss degree when the gradation skip does not exist. This is characterized in that the amount is determined, and has an effect that an optimal gamma correction without generating a false contour according to input image data can be automatically performed.
[0010]
Next, in the gamma correction device according to the second aspect, the low gradation part is divided into two parts, and the gradation distribution number XD of the low gradation side area of the low gradation part and the high gradation side Is calculated based on the number of gradation distributions YD in the region of (1), and the value of XD / YD is determined, and the amount of gamma correction is determined according to the value of the degree of gradation reduction. This has the effect of correcting the gradation loss such as that by gamma correction to make the image easier to see.
[0011]
(Embodiment 1)
Embodiments of the present invention described in claims 1 and 2 of the present invention will be described below with reference to FIGS.
[0012]
FIG. 1 is a block diagram of a gamma correction device showing the basic configuration of the present invention. In FIG. 1, reference numerals 1, 2, and 3 denote a histogram creation circuit for creating a histogram showing a gradation distribution from image input data R, G, and B. A tone skip detection circuit, 7, 8 and 9 are tone dropout degree detection circuits that detect the tone dropout level of each signal, and 10 is a tone dropout maximum value that detects the maximum tone dropout level of each signal. A detection circuit 11 is a γ correction coefficient calculation circuit that calculates a γ correction coefficient from the detection result of the gradation collapse degree and the gradation skip, and 12, 13 and 14 perform γ correction of each signal according to the γ correction coefficient. It is composed of a correction circuit and delay circuits 15, 16, 17 for timing adjustment.
[0013]
The image input data is 8-bit data for each of R, G, and B. First, histograms of the image input data R, G, and B are created by the histogram creation circuits 1, 2, and 3, and the presence or absence of a tone skip is detected by the tone skip detection circuits 4, 5, and 6 from the created histograms. I do.
[0014]
FIG. 2 shows a histogram indicating the state of skipped gradations, and the distribution is 0 at various gradation levels in a low gradation part. Normally, gradation skipping is easy to recognize in a low gradation portion up to the signal level 50 of the histogram, and the image quality is further deteriorated by performing γ correction in a state where gradation skipping is present. Is detected, no γ correction is performed. The number of consecutive tone levels at which the frequency is 0 in the histogram is determined by the tone expression performance of a printer or the like, and it is determined that tone skipping occurs and γ correction is not performed. That is, the number of continuous levels is reduced in a printer that can be noticed even with a small amount of gradation skipping, and is increased in other cases. In the embodiment, the number of continuous levels is set to 2, but usually takes a value of 1 to 3. If any of the input image data R signal, G signal, and B signal has a gradation jump, the information is sent to the γ correction coefficient calculation circuit 11, and as shown in FIG. That is, γ correction is not performed. The detection of the gradation skip in the low gradation portion up to the gradation level 50 is as described above. If the γ correction is performed in a state where the gradation skip exists in the low gradation portion, the image quality is further deteriorated. This is because false contours are generated and noise becomes particularly noticeable. Therefore, deterioration of image quality can be prevented by detecting gradation skipping in a low gradation part and determining whether or not to perform γ correction.
[0015]
Next, the degree of gradation collapse of each signal is detected by gradation collapse degree detection circuits 7, 8, and 9. The γ correction amount is determined based on the degree of the crush. 4 and 5 are explanatory diagrams for the method of calculating the degree of gradation loss. The low gradation portion is divided into two by gradation levels X1 and X2, and the number of gradation distributions X in the region on the low gradation side up to gradation level X1, and the number of gradation distributions in the region from gradation level X1 to X2. From Y, the value of the gradation collapse degree = X / Y is calculated. If X / Y ≧ 1, it indicates a state in which gradation loss has occurred, and if X / Y <1, it indicates a state in which gradation loss has not occurred. As an example of division, when the low gradation part is divided into two parts, as shown in FIG. 5, X3 = 10, X4 = 30, and X5 = 50 except for an area close to 0, and gradation division is detected. I do. The maximum gradation value detection circuit 10 selects the maximum value of the gradation loss degree of the image data R, G, B. Since the γ correction coefficient is determined based on this maximum value, even if any of the signals of the image data R, G, and B causes the gradation loss, the gradation loss can be corrected.
[0016]
Next, the procedure of the γ correction will be described with reference to FIG.
(1) Create a histogram of the image data R, G, B signals.
(2) For each of the image data R, G, and B signals, the number of distributions from gradation levels 0 to 30 is checked.
(4) It is checked whether any one of the image data R, G, and B signals has a continuous distribution of two gradation levels and has a distribution of 0, that is, a gradation skip. If there is a gradation skip, γ correction is not performed, that is, γ = 1. If there is no gradation skip, the next step is examined for gradation loss.
(5) For each of the image data R, G, and B signals, the number of gradation distributions XD from gradation levels X3 to X4 and the number of gradation distributions YD from gradation levels X4 to X5 are calculated.
(6) Detect the maximum value of the gradation loss degree XD / YD of the image data R, G, B signals.
(7) It is determined whether XD / YD is greater than 1, and if XD / YD is less than 1, γ correction is not performed, that is, γ = 1.
(8) If XD / YD is greater than 1, calculate a γ correction coefficient. That is, γ = −0.25 * (maximum value of XD / YD) +1.25 is calculated.
(9) It is determined whether γ is smaller than 0.5. If γ is larger than 0.5, γ correction is performed using the calculated γ value as it is.
(10) When the calculated value is smaller than 0.5, γ is corrected by setting γ = 0.5.
[0017]
As described above, in the case where the gradation skip has occurred, the γ correction coefficient calculation circuit 11 does not perform the γ correction by setting the γ correction coefficient = 1 regardless of the maximum value of the degree of gradation collapse. Further, only when the gradation skip does not occur, the γ correction coefficient corresponding to the maximum value of the gradation collapse degree is calculated. That is, when XD / YD <1, it is determined that there is no gradation loss, and the γ correction is not performed with the γ correction coefficient = 1. If XD / YD ≧ 1, it is determined that there is gradation loss, and the γ correction amount is determined in accordance with the degree of gradation collapse. If the degree of gradation collapse is large, the γ curve is greatly increased as shown in FIG. 3B, and if the degree of collapse is not so large, the γ curve is slightly raised as shown in FIG. 3C. . When the γ correction coefficient is determined in this way, the gradation in the high gradation part is slightly lost. However, in a backlight image that requires γ correction, correction is necessary from the low gradation part to the middle gradation part. , There is no concern about the gradation loss in the high gradation part. Also, there is no decrease in the contrast of a portion that requires γ correction.
[0018]
The gamma correction coefficients are input to gamma correction circuits 12, 13, and 14, respectively. The image input data R, G, and B are corrected to R ', G', and B ', respectively, by the gamma correction circuit. As a result, the same γ correction coefficient is applied to each signal, which has an excellent characteristic that the color balance is not lost.
[0019]
【The invention's effect】
As described above, according to the present invention, the optimum γ correction can be automatically performed on an image having no false contour by detecting the gradation skipping degree and the gradation loss degree.
[Brief description of the drawings]
FIG. 1 is a block diagram of a gamma correction device according to an embodiment of the present invention; FIG. 2 is a diagram for explaining gradation skipping in an embodiment of the present invention; FIG. 3 is a gamma correction according to an embodiment of the present invention; FIG. 4 is a diagram illustrating a coefficient. FIG. 4 is a diagram illustrating a histogram for explaining the calculation of the degree of gradation loss in the embodiment of the present invention. FIG. 5 is a diagram illustrating the calculation of the degree of gradation loss in the embodiment of the present invention. FIG. 6 is a diagram showing another histogram. FIG. 6 is a flowchart showing steps of a gamma correction device according to an embodiment of the present invention. FIG. 7 is a block diagram showing a gamma correction device in a conventional example.
1, 2, 3 Histogram creation circuit 4, 5, 6 Tone skip detection circuit 7, 8, 9 Tone loss degree detection circuit 10 Tone loss degree maximum value detection circuit 11 γ correction coefficient calculation circuit 12 R signal γ correction circuit 13 G signal gamma correction circuit 14 B signal gamma correction circuit 15, 16, 17 Delay circuit

Claims (2)

The input video signal R, G, and histogram creation means for creating a histogram that shows the tone distribution of the B signal, and the gradation jump detecting means for detecting a tone jump in the low gradation part from said histogram from said histogram A gradation crushing degree detecting means for detecting the gradation crushing degree of the low gradation part for each of the R, G, B signals; and detecting the gradation crushing degree of the low gradation part for each of the R, G, B signals, A maximum value of the degree of crushing, which is a maximum value of gradation crushing degree detecting means, a gamma correction coefficient calculating means for calculating a gamma correction coefficient based on the maximum value of the gradation skipping degree, and a gamma correction coefficient calculating means; R, G, and and a gamma correction unit for performing gamma correction for each B signal, when the gradation jump is detected by the tone jump detecting means as one of the gamma correction coefficients, tone jump is detected is When there is no gamma correction device characterized by having a gamma correction coefficient calculating means for calculating the gamma correction factor corresponding to the gradation loss degree maximum. The gradation loss degree detecting means divides the low gradation portion into two for each of the R, G, and B signals, and calculates the gradation distribution number XD of the low gradation region of the low gradation portion and the high gradation side. And calculating a value of the gradation collapse degree = XD / YD from the gradation distribution number YD of the region, and determining the gamma correction amount according to the value of the gamma correction coefficient by the gamma correction coefficient calculation means. The gamma correction device according to claim 1, wherein

Images