JPH0437977B2 - - Google Patents

Description

[Detailed description of the invention] This invention relates to a method for processing photographic prints.

Conventionally, when printing photographs, a method has been used to control the printing exposure of negative film for blue (B), green (G), and red (R) to a constant level.
This is derived from the empirical result that most photographed scenes have a constant hue and a constant brightness when their entire area is integrated.
If the subject differs from this standardized scene, controlling the print exposure to a constant value will result in extremely bad prints. For example, if you set up your printer to produce good prints for standard scenes (neutral), and then print a negative of a person in a red jacket under those conditions, the whole thing will turn gray, so the red Prints with the so-called cyan richness that is missing can be obtained.
Negative images of scenes in which the colors are biased toward one or two colors are called subdivision color effects, and conventionally, these had to be artificially controlled. Also, negatives with uneven brightness in scenes, such as negatives of people photographed in the snow, may produce poor prints if printed under conditions set to produce good prints from negatives of standardized scenes. is created. Negatives taken from such scenes are called subdivision density areas, and this also had to be artificially corrected.

In this way, a printing method that controls the amount of transmitted light over the entire area of the negative to be printed, that is, the Large Area Transmittance (Large Area Transmittance)
Printing methods that control density (hereinafter referred to simply as LATD) correct for 1) changes in LATD due to overexposure and underexposure and the influence of shooting light quality, producing good prints, but 2) sub-image If the density changes due to the effective color, a bad print will be produced, and even if the LATD value is detected, neither of the above 1) and 2) will result.
It is not possible to distinguish whether LATD changes or not. Therefore, it is assumed that LATD changes due to causes that occur more frequently, that is, cause 1), and LATD changes due to cause 1) are automatically controlled, and changes due to cause 2) are artificially corrected. means are adopted.

In recent years, efforts have been made to automate printing, and corrections for color aliasing have been automatically corrected to some extent by the adoption of lower word correction, but this has a negative effect on changes in photographic light quality, and the density aliasing has been reduced. For automatic detection, a method has been devised in which the density distribution is detected by optically scanning the negative surface, and the density of the subject is predicted based on the positional probability of the subject's presence and printed. However, it is still unsatisfactory. For such correction, it is necessary to accurately grasp the characteristic values of the negative film, and there is a strong demand for highly accurate detection of the film characteristic values.

In light of the above, an object of the present invention is to provide a method for printing photographs in a good manner.The purpose of the present invention is to provide a method for printing photographs in a good manner, by measuring the light of an original picture using a large metering dot size and a small dot size. The analysis is performed using the most suitable data according to the characteristics of the original picture.

This invention will be explained below.

This invention relates to a photo printing processing method in which an original photograph is divided into a large number of sections and photometered, and characteristic values obtained from the photometric values are used to determine the print exposure amount and printed. is divided into two types of sections, large and small, and densely measured with large photometering points of 0.5 mm square or larger, and 0.5
A small photometering point with a size of less than 1 mm square measures light densely regardless of the position of the large photometering point, and a skin color density characteristic value in the screen is determined from the photometric value of the small photometering point, and the large photometering point A characteristic value of partial area density in the screen is determined from the photometric value of , and when a skin color density characteristic value is detected, a printing exposure amount is determined from the characteristic value of the partial area density and the skin color density characteristic value,
If the flesh color density characteristic value is not detected, the printing exposure amount is determined from the characteristic value of the partial area density. This is because it is difficult to measure and distinguish the features of minute parts of an image with a large photometric point size, that is, a large section (for example, 1 to 2 mm square on a 135 mm negative film), and with a small photometering point size, that is, a small section ( For example, for a 135mm negative film (approximately 0.3mm square), the maximum density Dmax and the minimum density
This is based on the fact that it is inappropriate for measurements such as Dmin or for determining average density, and the photometering point size is selected in large and small sections depending on the type of characteristic value sought from the original photograph. Thereby, the characteristic values of the original image can be extracted under optimal conditions.

That is, FIGS. 1A and 1B show how the entire surface of the original photo 1 is divided equally into large sections 2 and equally divided into small sections 3, respectively.
By sequentially scanning the entire surface of the original photo 1 in the large section 2 and small section 3 as shown in the figure, and calculating the characteristic values obtained from each section, it is possible to accurately analyze the original photo 1. becomes. The optimum values for determining the characteristic values of the photographic original image 1 using large section 2 are average density, maximum density Dmax,
Minimum density Dmin, average density at the center of the screen, average density at the top half of the screen, average density at the bottom half of the screen, RGB
and YMC (yellow, magenta, cyan) and other hue areas. In addition, the optimum characteristic values to be determined by photometry using small section 3 are the average density of the skin-colored area (for example, as shown in JP-A-52-156624);
Histogram of absolute value of density difference between adjacent points
DBH=｜Di−D _i+1 | and its average value DB=Σ｜D _i
−D _i+1 |/n, spatial frequency distribution by Fourier transform, and the like. Note that Di indicates the concentration value in an arbitrary small section, and n indicates the number of measurement points.
The characteristic values obtained by photometry of photographic image 1 using large section 2 and small section 3 are used for scene classification and calculation of exposure amount, and for normal images, the exposure amount is determined by the characteristic value of large section 2. Do it like this. For small images of people or special images such as fluorescent light negatives and green back negatives, it is optimal to perform photometry in the small section 3 to obtain characteristic values. Also, from the characteristic values of subdivision 3, binarization or N
It is possible to create a valued image and to determine whether there is a low-density part (corresponding to estimation of hair) near the skin color extraction part.

Note that the scanning photometry positions of the small section 3 do not need to be adjacent to each other; for example, as shown in FIG.
As shown in 3A, photometry may be performed concentrically with the large section 2 at intervals, and it is not always necessary to perform photometry concentrically. And large section 2
The photometric point positions of the subdivision 3 and the photometric point positions of the small section 3 may have a corresponding relationship as shown in FIGS. 1A and B, or they may be photometrically measured independently. Furthermore, the photometering point of subdivision 3 is instead of the three colors of RGB,
A single color with a wide spectral distribution is also possible, and a flying spot scanner can also be used for photometry of a small area. The large section photometric value may be obtained by combining the small section photometric values.

Here, to explain the large section size and correction performance, the correction calculation formula for LATD is X=K ₁ +K ₂・Dmax+K ₃・Dmin+K ₄・D _LATD
...(1), and the rate of NG (No Good) is shown in Figure 2. Here, ±20% of the appropriate exposure amount
The above are unacceptable. Also, as a correction calculation formula: X=K ₁ +K ₂・Dmax+K ₃・Dmin+K ₄・D _LATD
+K ₅ [of (center - whole)] +K ₆ [of (upper half - lower half)] ...(2) can also be used (in Figure 2), but this (2)
If formula (1) is sufficient, it may be deleted. The photometric area is constant at 60% of the entire screen of a 135mm film, i.e. photometric area = section size x number of sections...(3).

As is clear from the characteristics shown in FIG. 2, a good large section size is 2 mm to 0.5 mm for a 135 mm size film, and a desirable section size for half size and 110 size films is proportional to the screen area. Also, the parcel size is 0.5mm for 135mm size film.
~0.1mm is good, and for half-size and 110-size films, a section size proportional to the screen area is preferred. However, small screen size films such as half size and 110 size film require lower quality than 135mm size film, so the large section size and small section size are not necessarily proportional to the area of 135mm size film. There's no need to make it smaller.

Next, an example of scanning an original photograph by photometry of the above-mentioned large section and small section will be described with reference to FIG. The photometry systems for large and small sections have exactly the same configuration, so here we will use the photometry system L for the large section.
I will explain about it.

The light emitted from the light source 10 is transmitted to the scanning mirror 11
After being reflected by L, the light passes through a lens 12L consisting of a condenser lens and a cylindrical lens, and then irradiates the film 20 that is being transported, for example, in the direction A in the figure. The light transmitted through the film 20 is further transmitted to the lens 1.
After passing through 3L, we reach dichroic mirror 14L,
The light reflected by the dichroic mirror 14L passes through a color separation filter (blue) 16L and is input to the line sensor 30L.
The light transmitted through L reaches a dichroic mirror 15L disposed further downstream, and the light reflected by this dichroic mirror 15L passes through a color separation filter (red) 17L and is input to a line sensor 31L. The light transmitted through the dichroic mirror 15L is input to the line sensor 32L via a color separation filter (green) 18L. Here, an opening member of approximately 1.0 mm square is linearly arranged at the input section of each of the line sensors 30L to 32L, and the input section of the photometric line sensors 30S to 32S forming one of the small sections. for,
Opening members, each approximately 0.3 mm square, are arranged in a linear manner. In addition, color separation filters 16L and 16
S is a filter that transmits blue light,
The color separation filters 17L and 17S transmit red light, and the color separation filter 18
L and 18S each transmit green light. Further, the mirrors 11L and 11S are each vibrated by a separate drive motor or the like so that the light spot irradiating the film 20 is scanned in a direction perpendicular to the transport direction A thereof.

In such a configuration, when the negative film 20 is conveyed in the arrow direction A shown in the figure, the transmitted light irradiated onto the negative film 20 is transmitted through each lens 13L.
and 13S, and the resulting images pass through dichroic mirrors 14L, 15L and 14S, 15S to line sensors 30L to 32L and 30, respectively.
It is input to S to 32S. Here, since the light irradiating the negative film 20 is scanned in a direction perpendicular to the transport direction A of the film 20 by the vibrations of the mirrors 11L and 11S, the line sensors 30L to 32L and 30S to 32S The screen is scanned as the negative film 20 is conveyed, and the entire image is detected in a large section (photometering system L) by the detection element of the line sensor through each aperture.
And photometry can be performed as a small section (photometry system S). In this way, the entire image on the negative film 20 can be photometered in the above-mentioned large sections and small sections.

In this invention, in addition to photometry in large sections, characteristic values of negative film images are measured in small sections.
diameter), it is possible to reliably extract
The possibility that the overlap between red and yellow will be detected as skin color is also reduced. Such points cannot be achieved by photometry of large areas. Additionally, in the past, it was difficult to distinguish between negatives photographed using fluorescent lights, as shown in the circle in Figure 4, and negatives photographed with low contrast, such as the one shown in the same figure.
Discrimination can be made by examining the absolute value of the density difference between adjacent photometric points in a small section. This is because the green back negative contains many minute contrast areas. Furthermore, when a photograph is taken in a garden or in front of plants in winter, there are many small high-contrast areas, but the contrast is difficult to detect using large-scale photometry. However, if the characteristic values are determined by photometry of a small area as in this invention, the lighting conditions and the background of the photograph can be predicted based on the contrast between light and dark in a minute area and the DB value, which is the average value of the above density difference. You will be able to do this. Furthermore, the photometric values of small sections are used to extract contour lines, which is done in normal image processing (for example, the Journal of the Institute of Electronics and Communication Engineers "VOL.59, No.
11 (1976), p. 1282) and image texture analysis (for example, Measurement and Control “VOL.16, No.2 (1977),
12 pages”), scene analysis in images (e.g. Pratt,
“Digital Image Processing”, Chapter 5, John
Wiley (1978)) can also be performed on a high-speed computer.

As described above, according to the printing processing method of the present invention, since the original photo is photometered in small sections and large sections, the original photo can be analyzed from a wide field of view, and the characteristics of the original photo can be accurately determined. can be grasped. Therefore, in photographic prints, by exposing using these photometric data, it is possible to realize better photographic prints.

[Brief explanation of drawings]

Figures 1A and B are diagrams showing the photometry of a large division and a small division, respectively, according to the present invention, Figure 2 is a characteristic diagram showing the NG rate of a large division size, and Figure 3 is a diagram showing the photometry of a large division and a small division used in this invention. FIG. 4, which is a block diagram illustrating the photometric system of a small section, is a histogram of a fluorescent light photographic negative and a greenback negative. 1...Original photograph, 2...Large section, 3,3A...
Small section, 10... Light source, 11L, 11S... Mirror, 12L, 12S, 13L, 13S... Lens, 14L, 14S, 15L, 15S... Dichroic mirror, 16L to 18L, 16S, 18
S...Color separation filter, 30L to 32L, 30
S, 32S...Line sensor.

Claims (1)

[Claims] 1 Divide the original photograph into many sections and measure the light.
In a photographic printing processing method in which the printing exposure amount is determined using characteristic values obtained from the photometric values, the entire surface of the photographic original is divided into two types of sections, large and small, and 0.5
In addition to densely measuring light with a large-sized photometering point with a size of mm square or larger, photometry is also carried out densely with a small-sized photometering point with a size of less than 0.5 mm square, regardless of the position of the large-sized photometering point, and photometry of the small-sized photometering point is performed. The characteristic value of the skin color density in the screen is determined from the value, and the characteristic value of the partial area density in the screen is determined from the photometric value of the large size photometric point, and if the characteristic value of the skin color density is detected, the characteristic value of the partial area density is determined. and the skin color density characteristic value, and if the skin color density characteristic value is not detected, the printing exposure amount is determined from the characteristic value of the partial area density. Processing method.