JP3112475B2 - Image processing method - Google Patents

Description

Description: TECHNICAL FIELD The present invention relates to an image processing method for performing gradation conversion on an input color image.

[Prior Art] Generally, the black level of a video signal differs depending on the video device. Further, the white level also differs depending on the video device, and also differs depending on the shooting conditions. Even if there is a difference between these levels, since the brightest part is recognized as white and the darkest part is recognized as black on the television monitor, unnaturalness is not felt.

However, a problem arises when printing out this video signal. For example, if a video signal with a low white level and a high black level is printed out, the white part will appear dark and black will not be printed as black, so there is no overall reduction. I feel.

As one method for eliminating such a difference between a television monitor and a printout, a method of changing a gradation conversion curve of input data on a printer side is known. This is the fifth
Shown in the figure. In this way, both the low-luminance part and the bright part are slightly rubbed, so that they are pure white (no printing).
And printed in black. Further, since a curve near the intermediate value also rises, a print image with a reduced tension can be obtained.

The same applies when the RGB signal is simply inverted to a density value (CMY) signal (see FIG. 6). That is, when the CMY value is large according to the gradation conversion curve, the value is further increased, and when the CMY value is small, the value is further decreased. Therefore, the input values are converted as shown in FIG. 7, and as a result, the saturation of the print image is increased.

For the reasons described above, it is common to perform correction by gradation conversion in a video printer.

[Problem to be Solved by the Invention] In the above-described conventional gradation conversion processing, when an input image has values distributed around an intermediate value, a vivid print can be obtained without any problem.

However, a problem occurs when the input image is shifted toward a bright side or toward a dark side. Here, the shift to the bright side is when overexposure occurs due to a shooting error of a video input device (still video camera, video movie camera, or the like). At this time,
In spite of the overall brightness, the white portion is destroyed due to further gradation conversion during printing.
In this case, the crushed white area occupies a considerably large area, so that the user feels very unnatural.

Contrary to this, when approaching the dark side, it occurs when taking a flash with a still video camera. In flash photography using a still video camera, even if a target subject is photographed with appropriate brightness, the flash light hardly reaches the background of the subject, so that the subject is photographed dark. Therefore, if printing is performed by the above-described gradation conversion in this state, the background portion will be blackened. This problem exists even when gradation conversion is not performed for the following reason.

This is the lens of the brightness reproduced on the print,
It is due to the fact that it is much narrower than the lens with the brightness reproduced on the TV monitor. As a result, even if the background portion of the subject is identified on the television monitor, there is a problem that it is printed uniformly black on the print and cannot be determined.

Therefore, an object of the present invention is to provide an image processing method that performs appropriate gradation conversion on a flash image and an overexposed image in view of the above points, so that a good output image can be obtained. is there.

[Means for Solving the Problems] In order to achieve the above object, an image processing method according to the present invention includes the steps of: inputting a signal indicating the brightness of an input color image;
Counting the number of pixels in a highlight portion and a dark portion, and determining the number of pixels in the highlight portion and the dark portion to determine whether the input image is a flash image or an overexposed image. Comparing the number of pixels with a predetermined value set according to the number of input pixels,
Setting a tone conversion condition in accordance with the result of the comparison; and performing a tone process on the input color image using the set tone conversion condition.

[Embodiment] In an embodiment of the present invention, a method of preparing several types of gradation conversion curves and automatically selecting the input image when selecting the gradation conversion curve according to the situation of the input image, and automatically selecting the curve is used. Used. More specifically, a video signal is sampled and input, and if there is a number N _H or more of pixels having a luminance equal to or more than a certain threshold value T _H , it is determined that the image was captured with overexposure. Then, a gradation conversion curve (for example, as shown in FIG. 2) corresponding thereto is selected. As a result, it is possible to obtain a print in which a portion that flies white is improved as compared with the related art.

Conversely, if there is few N _L or more pixels in the following threshold value T _L with brightness, which is determined to have been flash photography. Then, a gradation conversion curve (for example, as shown in FIG. 2) corresponding thereto is selected. As a result, the background portion, which has conventionally been uniformly darkened, is improved.

If none of the above cases, it is determined that the image is a normal image, and a conventional gradation conversion curve (for example, in FIG. 2) is selected.

When sampling and inputting a video signal,
It is not necessary to sample all pixels. That is,
It is sufficient to perform sampling by thinning out the pixels to be printed every several pixels in the vertical and horizontal directions.

In addition, the above-described determination can be performed by a microcomputer on the printer side, and can be performed within a few seconds, for example, when a user presses a print button, paper is inserted, and a print start state is started.

That is, in implementing this embodiment, there are no disadvantages such as a slow print time, and no additional hardware is required.

Hereinafter, an embodiment of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram showing one embodiment of the present invention. Here, 2 is a Y / C separation circuit, 4 is an input video signal switching circuit, 6 is a decoder, 8 is an A / D converter, 10 is a memory controller, 12 is an image memory, 14 is a D / A converter, and 16 is an encoder. , 18 is a Y / C synthesis circuit, 20 is Y / color difference /
RGB conversion circuit, 22 is a changeover switch, 24 is an A / D converter,
26 is a print controller, 28 is a head driver, 30 is a head, 32 is a print unit, 34 is a body switch, 36 is a CPU (system controller), 38 is RAM, 40 is line memory, 4
2 is the ROM for gradation conversion curve, 44 is the ROM for pulse data, 46
Is a ROM for luminance / density conversion.

The input video signal is Y / C separated by a Y / C separation circuit 2. The S / Video signal and the Y / C separated signal are selected by the input video signal switching circuit 4, pass through the video decoder 6, and become Y, R−Y, and B−Y signals. Then, the signal at the time when the user presses the memory button (not shown) is taken into the image memory 12.

 Printing is performed as follows.

The Y, RY and BY signals are read from the image memory 12 and
R, G, R, G, and B signals from the Y, R-Y, and B-Y signals by the color difference / RGB conversion circuit 20
Converted to B signal. Generally, in a sublimation type thermal transfer printer, printing is performed in a sequential manner in the order of yellow (Y), magenta (M), and cyan (C).

So, first, the blue signal is selected by switch 22,
A / D conversion is performed by the A / D converter 24.

Next, gradation conversion is performed with reference to the gradation conversion look-up table (ROM). Then, the luminance / density conversion ROM 46 is read in order to convert the luminance blue to the density yellow.

Thereafter, in order to obtain pulse width data to be applied to the head 30, the pulse width conversion look-up table (ROM) 44 is referred to. At this time, the current temperature of the head is determined in consideration of the temperature information.

This pulse width data is taken into the line memory 40,
When one line of data is accumulated, the head driver
28 drives the head 30 to perform printing. This operation is repeated for one screen, and the yellow printing is completed.

Next, A / D conversion of the green signal is performed and printing is performed in the same manner, and printing of magenta is completed. Finally, the red traffic light
A / D conversion is performed and printing is performed in the same manner, and printing of cyan is completed. This completes all printing.

Next, an operation of determining overexposure and flash photography in the present embodiment will be described.

When the user presses the button of the main body side switch 34, the mechanical system starts a paper feeding operation. During that time, the luminance data Y is taken into the line memory 40 via the printer controller 26.
The Y data are read from the CPU 36, first, it is compared with overexposure threshold value T _H. And if T _H above, "1" is written to the RAM area of the CPU 36 (part of RAM 38).

Next, a comparison is made with the flash photographing threshold value _TL ( _TL <
T _H ). If it is equal to or less than _TL , "1" is written to another RAM area of the CPU 38. By repeating this, the Y data becomes _TH
If it is equal to or greater than or equal to or less than _TL , the contents of the corresponding RAM area are counted up. It is not necessary to perform the above processing on all the sampled Y data, and it is also possible to appropriately thin out (for example, sampling every four pixels both vertically and horizontally).

When the comparison is performed up to the last Y data in this way, the contents of the RAM area are compared with certain numbers N _H and N _L. That is, the contents of the overexposure RAM are compared with a certain number _NH .
In the present embodiment, _NH is set to の of the number of pixels required for comparison.
If it is larger than _NH, it is determined that the image is overexposed, and the curve in FIG. 2 is selected. FIG. 3 (a) shows an example of the histogram in this case, and FIG. 3 (b) shows the shape of the histogram when converted via the curve in FIG.

The content of the flash photography RAM is compared with a certain number _NL . In this embodiment, _NL is set to 1/2 of the number of pixels required for comparison. If the result is larger than N _L , it is determined that the image is obtained by flash photography, and the curve in FIG. 2 is selected. FIG. 3 (a) shows an example of the histogram in this case, and FIG. 3 (b) shows the shape of the histogram when converted via the curve shown in FIG.

In this way, both the overexposed image and the flash image are converted into improved image data. If neither of the above matches, a normal conversion curve (FIG. 2) is selected.

The above process is performed for several seconds until the paper is fed and the printing is started. Then, the above-described yellow, magenta, and cyan printings are performed.
Since the B, G, and R signals are converted by the selected gradation conversion curve, an improved print output is obtained.

In other embodiments above embodiment, as the gradation conversion curve has dealt with the case in the present 3 for normal flash photography for overexposure, several stages of T _L and T _H Alternatively,, N _H
By providing several levels of N and N _L , it is possible to select more gradation conversion curves and to perform more appropriate correction.

[Effects of the Invention] As described above, according to the present invention, appropriate gradation conversion can be performed on a flash image and an overexposed image without the user being particularly conscious, and the flash image and the overexposed image can be converted. An image can be reproduced well.

[Brief description of the drawings]

FIG. 1 is a block diagram showing an embodiment of the present invention, FIGS. 2 to 4 are diagrams for explaining the operation of this embodiment, and FIG. 5 is a conventionally known gradation conversion process. FIG. 6 is a diagram when density conversion data is obtained by simply inverting a gradation conversion curve, and FIG. 7 is an explanatory diagram showing an effect of correction by the gradation conversion curve. 2 ... Y / C separation circuit, 4 ... Input video signal switching circuit, 6 ... Decoder, 8 ... A / D converter, 10 ... Memory controller, 12 ... Image memory, 14 ... D / A converter , 16 ... Encoder, 18 ... Y / C synthesis circuit, 20 ... Y / color difference / RGB conversion circuit, 22 ... Changeover switch, 24 ... A / D converter, 26 ... Print controller, 28 ... Head Driver, 30 head, 32 print unit, 34 body switch, 36 CPU (system controller), 38 RAM, 40 line memory, 42 ROM for gradation conversion curve, 44 …… ROM for pulse data, 46 …… ROM for brightness / density conversion.

Claims (2)

(57) [Claims] 1. A step of inputting a signal indicating the brightness of an input color image, counting the number of pixels in a highlight portion and a dark portion, and determining whether the input image is a flash image or an overexposed image. Comparing the number of pixels of the highlight portion and the number of pixels of the dark portion with a predetermined value set according to the number of input pixels to determine whether the tone conversion is performed according to a result of the comparison. An image processing method, comprising: setting a condition; and performing a gradation process on the input color image using the set gradation conversion condition. 2. The image processing method according to claim 1, wherein a gradation conversion curve in which a gradation width of a dark portion is wider than a gradation conversion curve of an overexposed image is used for a flash image. Method.