Controlling Method for a Photographic System

In the accompanying drawings:

FIGS. 1 and 2 are graphs for explanation of the prior art normalization;

FIGS. 3 through 5 are graphs which respectively describe one example of this invention control method;

FIG. 6 is a flow chart to obtain average characteristics of a film according to this invention;

FIG. 7 is a flow chart to show one example of addition and initialization according to this invention;

FIGS. 8A and 8B are explanatory graphs of application of this invention method to the determination of photographic printing conditions;

FIG. 9 is a block diagram which shows one example of a device to realize this invention normalization method;

FIGS. 10A through 10C are graphs to describe normalization according to this invention;

FIG. 11 is a block diagram of another embodiment of this invention;

FIG. 12 is an explanatory graph to show how to determine density divisions;

FIG. 13 is a block diagram which shows a device to realize still another embodiment of the normalization method according to this invention;

FIGS. 14 and 15 are graphs which explain the normalization; and

FIG. 16 is a block diagram to show still another embodiment of the normalization according to this invention.

This invention attempts to control a photographic system which is subject to chronological changes by sequentially storing the photographic characteristics of the system, using the same as a chronological function to modify the importance of each characteristic to obtain a weighted average photographic characteristic. More particularly, positive weights are added to the photographic characteristics of predetermined time (e.g. one day or ten days) or a predetermined amount (e.g. 1,000 frames) consecutively in a manner that the weight becomes smaller as the data goes retrogressively afar from the current time point, and the control is conducted based on the weighted moving average photographic characteristics. The photographic characteristics of a predetermined time or amount may be determined as an average of individual photographic characteristics within a predetermined scope (e.g. time or amount), and that may be stored to obtain the weighted average photographic characteristics. Weight may be varied for each storage. The photographic characteristics of a predetermined time or amount may be weighted to sequentially renew the average or sum thereof and to change weight for each time as shown in FIG. 3. In the case shown in the figure, the weighted average up to M is multiplied with a weight coefficient W (W <1.0) and added to the average of (M +1) to renew the weight.

Alternatively, as shown in FIG. 4, a stored number N of the photographic characteristics is set in relation to time within the scope not above the limit N.sub.u and not below the limit N.sub.l and when the data number has reached the upper limit N.sub.u, it is transformed to the lower limit N.sub.L to start from the value anew. If the sum and the data number are stored, when the data number has reached the upper limit N.sub.u, the process starts anew from the value obtained by multiplying the sum and the data number with n (n<1 and n.noteq.0). Alternatively, an average value may be stored in place of the sum. More particularly, if the average A, the data number is N and the photographic characteristics at the current time point is A.sub.o, the equation below holds in storage;

and when the number N has reached the upper limit N.sub.u, as N=N.sub.l, the process starts again anew.

It is assumed that N.sub.L =N.sub.u .multidot.n, and n is a positive number smaller than 1 (for instance 1/2, 2/3. . . ).

FIG. 5 shows another method to determine the weighting coefficient. In this weighted moving average method, the data in the past is weighted relatively less than the data closer to the current time point to estimate the current situation.

FIG. 6 shows an example to obtain the average characteristic of film sheets. A sheet of film is photometrically measured (1) by an image sensor and so on, the type of the film is classified (2) as (A, B, .....N) based on the class information and the photographic characteristics Dare stored. When the stored number N is smaller than 2,000, it is operated and processed (3) with the average D.sub.x obtained from the stored data on all the film sheets, and when the data number is larger than 2,000, the stored average D.sub.x is not used, but the average characteristics obtained from the stored data on individual film sheets is sued for the control (4) for photographic printing. More specifically, if 100 film rolls are processed per day for one type of film (in other words, ca. 2,000 data/day), and if the range is limited at 4,000 at the ceiling and at 20,000 at the bottom, the weight is modified once in a 10-day cycle. The upper and bottom limits are preferably changed for each type of the film rolls depending on the amount of process. The changes in film development and seasonal variation are sequentially traced to detect the amount thereof and control is conducted based thereon. In the control based on the average of a large number of data, the contribution of a data of the 1,000th frame to the control differs from that of a data of the 100,000th frame. Although the precision as an average of the data of 100,000 frames is higher, it is contradictorily less efficient as the data to estimate the current conditions. In other words, although it takes a long time to store a large amount of data, the farther past the data concerns, the less it reflects the current situation. This invention solved such a contradictory problem by using a weighted moving average.

In FIG. 6 in the control of the printing conditions, for example, the average characteristics is expressed in the average LATD (Large Area Tansmittance Density) and the operational processing is conducted on the difference between the reference LATD and an obtained averag LATD. It represents that the developing characteristics deviate from the reference value by the LATD difference and the printing conditions are corrected or adjusted to eliminate such a deviation. The operational processing needed for normalization of a characteristic curve which is variable depending on the type of film is the calculation of normalization conditions. If the data number is 2,000 or less, it may use the initial value which has been stored in advance and the system may include initial input means therefor.

FIG. 7 shows a flow chart to describe an example of addition and initialization.

More particularly, the data Di'on each frame is obtained by the photometric measuring with an image sensor and so on (Step S1); the density D'is added and the number N is added (Step S2); and the above measuring steps are repeated for one roll of the film (Step S3). After such a photometric measurement is conducted for a predetermined number times, it is processed by addition of the average A and the number N stored in a memory (Step S4). If the number N has not reached a predetermined upper limit N.sub.u, the average and the number thereof are stored in a memory as they are (Steps S5 and S8), and if it has reached the upper limit N.sub.u, the number is returned and stored as the bottom limit N.sub.L (Steps S6 and S8). Above steps are repeated until the suspension of the measurement (Step S7). The data stored in the memory should be erased when the developing agent is replaced, the type of film is switched to another type, or the photographic system or the printer fails. it is desirable to include a means which can erase the stored data or return the data to an initial value with an external input.

This invention method may be applied in various manners other than above in the photographic printing condition determination; for instance, it can be utilized in determination of a coordinate point, a coordinate region or correction amount on a color coordinate system of a film which has been exposed with tungsten light. It is also applicable to the determination of the tendency of density correction or color correction to obtain the most suitable conditions. It may be aplied to the detection and correction of the chronological deterioration or the various changes in measurement device of the image, display unit or output device. FIG. 8A shows the clipping of data in relation to time while FIG. 8B exemplifies the weighting coefficients of the past data.

Although the photographic characteristics are variables of time, since they are weighted with the data up until the current time point by estimating the current conditions in this invention method, suitable and precise control and administration thus becomes possible.

FIG. 9 is a block diagram of the system which realizes this invention normalization method wherein a color negative film as an original film is photometrically measured (10) in the three primary colors of RGB with photodiodes or an image sensor, the characteristics 15 (e.g. average density, the maximum density, the minimum density) for exposure control are obtained from the density values due to the data obtained in the measurement, the exposure amunt is calculated (16) with the characteristics 15 in accordance with a predetermined formula, and a photographic printer is controlled (17) in exposure with the obtained exposure amount. Simultaneously, the characteristics 11 (such as average density on the whole or portions of a frame, the maximum or the minimum density, density histogram or hues, etc.) on a frame are obtained, and the characteristics 11 are stored in a memory 14. From the data stored in the memory 14, the correction coefficients (or constants) 13 are obtained to correct (12) the density values obtained from the measurement 10 and the characteristics 15 for the exposure control are calculated based upon the result of the above correction.

The characteristics 11 obtained from the photometric measuring 10 are compared with predetermined density divisional values, the average of frames having the characteristics larger than the division are expressed as A.sub.j while the average data of frames having the characteristics smaller than it are expressed as B.sub.j. As j=1, 2, 3 or R, G, B, the correction coefficient 13 to normalize the measured data is obtained by A.sub.j and B.sub.j or combination thereof. For instance, if it is assumed that the measured data is Dij and the correction value is D'.sub.ij, the correction coefficients for gradation and the correciton coefficients for density balance may be respectively obtained by a combination of A.sub.j and B.sub.j. When normalization is conducted based on the data G out of the measured data on the three primary colors, it may be obtained by the following calculation. ##EQU1## when the measured data on RGB should be normalized based upon a predetermined film characteristic curve, it may be calculated according to the following equation (2). ##EQU2##

K in the former equation (1) may be B.sub.2 or it may be calculated based on B.sub.1 or A.sub.1. S.sub.j and G.sub.j in the above equation (2) are reference values. Correction with B.sub.1, B.sub.2 and B.sub.3 may be conducted after the gradient on D.sub.ij is normalized in the above equations (1) and (2). In place of D.sub.i1, D.sub.i2 and D.sub.i3, a color difference value comprising a combination of two or three colors thereof may be obtained. The normalization method of this invention is by no means limited to the above equations (1) and (2). For instance, a predetermined and present value may be used to replace the gradation correction value in the equation (1).

FIGS. 10A through 10C show the states of RGB-normalization of the meaured data for G and R based on G according to the equation (1). FIG.10A illustrates actual data obtained from the photometric mesurement. Gains are corrected in order to make B.sub.1 =B.sub.2 by means of the equations R'=R-B.sub.1 and G'=G-B.sub.2 (FIG.10B). The data are further normalized by .gamma.-correction to achieve the relation R'=(R-B.sub.1).multidot.{(A.sub.2 B.sub.2)/(A.sub.1 -B.sub.1)}+B.sub.2 and G"=G. The state illustrated in FIG. 10C is thus obtained. The same object can be reproduced at the same density and in the same color by transforming the density values at the normalization conditions obtained respectively for each type of the film. This enables the exposure amount to be calculated at the same and constant conditions irrespective of variation of the film types or the processing changes. Only one kind software will suffice for the arithmetic operation on the exposure control.

The characteristics 15 for the exposure control may be obtained by correcting the density values 12 obtained from the measured data with the correction coefficients 13. A frame average density is used as the frame characteristics or DL.sub.j which will be calculated in accordance with the following equation (3). ##EQU3## In determining the printing conditions, if a section for setting the film characteristics is separated from a prior art system, the exposure may be controlled with the density value normalized by this invention method to automatically determine and administer the printing conditions.

FIG. 11 shows another embodiment of the normalization method according to this invention wherein the density values obtained from the photometric measurement 20 is normalized (21) thereby to obtain the characteristic 22 on a frame. The exposure amount 23 is calculated in correspondence to the characteristics 22 for the exposure control 24. The normalization steps comprise transforming the characteristic 22 into a nonnormalized frame characteristics 25, storing the transformed data in a memory 26 to obtain the normalized coefficients (or constants) 27, normalizing the density data obtained by the photometric measuring 20 in accordance with the above values and correcting the transformed data 25. The steps of coverting the measured data into the density data by logarithmic transformation and of normalizing the density data can be conducted simultaneously.

The number of the density divisions mentioned above may be more than two. If it is two, the averages of the characteristics can be obtained in the number of three to form two pairs of the correction coefficients each for the low density zone and the high density zone. Moreover, the divisions may be set at X.sub.1 through X.sub.4 as shown in FIG. 12 to make the density regions overlap continuously to be divided into the regions less than X.sub.2, from X.sub.1 to X.sub.4, and more than X.sub.3. By setting a large number of density divisions, it becomes possible to estimate the shapes of the film characteristic curve, from which overexposed portion or underexposed portion can be controlled suitably. A density division may be a predetermined constant or an average of the characteristics obtained from the measured data, and preferaby set at a value to make the nhumber of data above and below the division substantially equal in number. It is desirable to be able to select or correct the data by eliminating abnormal values (such as low density, high density or data extremely out of the color balance) or setting a limit to obviate them so as to enable convergence of a small number of data into an average. An upper and a lower limits may be provided for stored data numbers (frame numbers (so that when it reaches the upper limit, average processing of the data should be started again from the beginning, and until it reaches the lower limit, the initially set value may be used rather than corrected value or the average of all the negative films including all types of the film may be used as a correction value. The correction value may be obtained for each type of the photographic materials (or onr as chosen to cover several types), averaging of data may be stored so that averaging process may be conducted once an order or once a day. Alternatively, the sum of characteristics may be stored in place of averages. The averaging process may be taken in this case at the time of switching printer channels at the beginning of an order.

FIG. 13 is a block diagram to show still another embodiment of the device to realize this invention normalization method in correspondence with the block diagram shown in FIG. 9. A conversion talbe 18 is prepared for normalization by using the data stored in the memory 14, the data in density obtained from the photometric measuring 10 are corrected (12), and the characteristics 15 for the exposure control are obtained based upon the correction. The photometric measuring 10 may be conducted twice; one each for calculation of the characteristics 11 on a frame and for correction (12) of the density values. Or the density values may be stored in a memory and corrected the stored data in the memory.

The effect of storage of frame characteristics at the memory 14 will now be described. As is widely known, almost all the frames of film are gray with slightly yellow greenish tint. It is possible to precisely produce the color by obtaining an average in LATD of a large number of film rolls. The result of experiments revealed, however, the values not necessarily in LATD but of other frame characteristics. Moreover, it was revealed that even if the area is divided into two; i.e. the densities more than the average and the densities less than the average, and the average is obtained once each for two frame characteristics, they show the same color. From such a finding, it was deducted that the film density may be normalized with the average of the frame characteristics of the two density areas. Frame characteristics are added (accumulated) at the data of the memory 14 and an average is obtained at the time of preparing the conversion table 18.

The conversion table 18 may be prepared as follows. The characteristics 11 obtained from the photometric measuring 10 are compared with the density division value which is preset, the accumulated data on a large number of frames having the characteristics larger than the value are designated as A.sub.j while those having the characteristics smaller than it are designated as B.sub.j. Using the relationship that j=1, 2, 3 or R, G, B, A.sub.j, B.sub.j or a combination thereof, the conversion table 18 may be prepared for data normalization. For example, if the measured data is D.sub.ij (i =number of measuring spots) and the corrected value as D'.sub.ij, a value for the optimum density balance and the gradation correction will be obtained by combining A.sub.j and B.sub.j. When the conversion table 18 is prepared for the normalization based on G out of the measured data on three primary colors, the values of D'.sub.i1 and D'.sub.i3 obtained by calculating D.sub.i1 and D.sub.i3 for "0" through "255" by the above formula (1) should be stored in the conversion table 18. When the conversion table 18 is prepared for normalizing the data on RGB based on a predetermined characteristic curve of the film, the above formula (2) should be relied.

As FIG. 14 shows, a curve is plotted for the average obtained from the stored data of R and B by referring to the reference or the average obtained from the stored data of G, from which is prepared a conversion table to transform the measured data on R and B into the measured data on G for normalization. The measured data on G do not have to be transformed by the conversion table, but are used as they are. By these normalization processing, the same object can be imaged on a negative film irrespective of the types of the photographic materials or developing processes.

The reference may be the measured data of (R+G+B)/3 as shown in FIG. 15. It may also be the data of (0.3X R+0.6X G+0.1X B )/10. Two or more of the RGB measured data may be combined suitably to be used as the reference. The cross marks in FIGS. 14 and 15 represent respective averages which are obtained by shifting R and B or R, G and B by a predetermined amount for the relationship between the color of the average and neutral color in the frame characteristics. The relation between the above two is not limited to the above, but may be otherwise included in calculation.

The conversion table 18 may be prepared in advance in the form of a table by calculating reference values for "0" through "255". In case the density division value is set at one point, the stored data larger or smaller than the point are obtained. This is a linear transformation. For instance, when a large number of measured data should be transformed, it may take lots of time if one of the equations of linear transformation (1) and (2) is to be applied to each of the measured data for two or three colors. Time can be shortened remarkably by using a conversion table. When the density division value is two or more in number, the transformation is nonlinear, and the conversion table may be prepared by linear interpolation or polyominal interpolation between the stored data. It is advantageous to use the conversion table rather than using a transformation formula in the speed of operation or degree of freedom allowable to non-linearity type.

Such a conversion table may be prepared at the time when a type of film is switched to another (channel switching), for one film roll, or for a predetermined amount of the stored data. It may also be prepared at the start of a work day for the channels to be used for the data based on the data accumulated by the preceding day. A magnification n (n= product by multiplication with 0.5-2.0) for each type of the film may be used as the reference value. The value of n is determined in advance. For example, n is set at 1.0 for the film of ASA 100 made by a manufacturer A while it is set at 1.12 for the film of ASA 100 by the manufacturer B. This enables elimination of the influence by the gradation of the film (.gamma.-value). Further, a conversion table may be prepared by plotting "the data before conversion" -"reference value" at the vertical axis.

The density values 12 obtained from the measured data are corrected by referring to thus mae convesion table 18 to obtain the characteristics 15 for the exposure control. For obtaining the average density, a frame is divided into 12 cells each of which comprises the measurement points of 6.times.6=36. The average of a cell may be used as an average density data for storage. When the amount of data is small, this method is preferable as it achieves a higher precision over a wider density area than the case where a whole frame is processed as a single data. If the whole frame is treated as one data, it can not have a high density as it has been averaged and the number of densities is small. Each of the measured point may be stored in a memory, but this is not very effective as the averages do not become uniform even for a large number of data.

FIG. 16 shows another embodiment of this invention in correspondence with the one shown in FIG. 11. A conversion table 28 is prepared for normalization by storing in a memory 26 the data for transformation into frame characteristics before normalization. Density values of the photometric measuring 20 are normalized (21) in a manner mentioned above in accordance with the conversion table 28 and the converted data 25 are corrected at the same time.

The above statement was given in relation to the normalization process for the measured data or density data, but the process may naturally be applied for the frame characteristics or the exposure conrol characteristics. This invention method is remarkable in processing speed as respective meaured points do not need to be transformed.

The frame characteristics to be used for averaging may be the average density(s) of a whole frame or portions of a frame, weighted average, average density at selected point(s), the mean value between the maximum and the minimum densities, the average density of a main object or of background or a combination thereof. The abnormality or deviation detected at the development processing or the photometric meaurement device from the changes in averages or normalization coefficients may be displayed or outputted as an alarm to automatically correct or replace the developing agents and/or modify the printing conditions. Moreover, the difference in averages between photographic materials is used to detect the difference in the printing conditions between the photo-sensitive materials to automatically correct and control the exposure amount. Although description is given to the case of a negative film in the above statement, other types of film such as the reversal film or printed photographic paper may be used in this invention. This invention method is applicable to other types of copying materials (thermal, magnetic, or other types) or to an image display device.

This invention normalization method enables automatic normalization at a high speed for various types of film to be printed withput need for any reference film. This invention method can be applied to non-linear type of normalization to precisely conduct normalization for low density portions. This invention method further makes correct judgement of colors in objects imaged with an artificial light which often causes the under-exposed images to thereby enable highly precise control over the exposure. This invention method is further applicable to automatic control of variation in the developing process conditions, the emulsion numbers, the seasons or the devices.

It should be understood that may modifications and adaptations of the invention will become apparent to those skilled in the art and it is intended to encompass such obvious modifications and changes in the scope of the claims appended hereto.