GB2154827A - Method for correcting picture signals in image input equipment - Google Patents

Abstract

In deriving picture signals from a multi-element sensor (1), instead of conventionally obtaining the ratios of picture signals to (white) reference signals, the dark signal of each element in turn is held in a D.C. clamping circuit (3), so that it can be subtracted from the element signal which is then stored in holding circuit (4). The corrected reference signals are transferred therefrom to RAM (6), and subsequently read out so that the ratio of the corrected picture signal from (4) to the corrected reference signal for the same element is derived in dividing circuit (2). <IMAGE>

Description

SPECIFICATION Method for correcting picture signals in picture input equipment This invention relates to a method for correct ing picture signals in a picture input equip ment so that an image can be read in as close as possible to an original by means of a photosensor such as photodiode array or CCD in which a plurality of photoelectric transducers are arranged.

In a conventional correction method of picture signals obtained by a photosensor, the correction is effected, as disclosed for example in Japanese Patent Publication No.

19184/1983, Japanese Utility Model Laidopen No. 19566/1983, Japanese Patent Laid-open No. 1 14310/1g74, etc., etc., by stor- ing information data on variations in sensitivity of individual elements of the photosensor and non-uniformity in illuminance of the lighting system and optical system and then adjusting the gains of signals, which have been obtained from an actual picture, on the basis of the data.

According to the above mentioned conventional method, dark current fractions inherent to respective elements of the photosensor are not taken into consideration. Therefore, the prior art method is unable to fully perform correction over the entire tonal range, especially in the dark range although correction over such ranges is necessary in the fields of printing plate fabrication, printing and the like.

On the other hand, a method taking the correction of dark currents into consideration is disclosed in Japanese Patent Laid-open No.

139139/1983, in which output values (dark current levels) of individual elements when they are shielded from light are also stored whereby to perform a correcting operation.

This method however requires a complex circuit, which leads to a cost-up. Furthermore, two operations are required, one in the bright and the other in the dark, in order to input correcting reference signals. The method is accompanied by a further drawback that it ends to develop correction errors if the dark current levels vary due to a temperature change or the like in the course of read-in of an actual picture.

An object of this invention is therefore to provide a method for effecting with ease and high accuracy correction of picture signals in a picture input equipment making use of a multi-elements photosensor which is used in the field of printing plate fabrication, printing or the like where a high degree of fidelity is required relative to originals, whereby to avoid influence due to inherent variations in sensitivity among individual elements of a photosensor, variations of dark currents, or changes in illuminance of the optical and lighting systems.

The method of this invention features that in a picture input equipment making use of a multi-elements photosensor, prior to performing an operation on (actual picture signals)/ (reference signals) in accordance with a conventional correction method of picture signals, dark currents from the individual elements of the photosensor are first corrected respectively by representative values of dark output values of the corresponding elements upon both reading in real picture signals and reading in reference signals and the above correcting operation is then performed based on the thus-corrected values.

In one aspect of this invention, there is thus provided a method for correcting picture signals in a picture input equipment adapted to read in an actual picture by means of a multielements photosensor, which method comprises: reading in a picture of a uniform density prior to the read-in of the real picture; encoding values which have respectively been obtained by subtracting dark output values of the individual elements of the photosensor from the values read above by the respective elements, and storing the thusencoded values as correcting reference signals for their corresponding elements; and upon reading in the real picture, reading out the correcting reference signals in synchronization with the read-in of the real picture so as to correct values, which have been obtained by subtracting the dark output values from the thus-read values of the real picture, on the basis of the above read-out correcting reference signals.

According to the method of this invention, picture signals can be corrected, with good accuracy and by a simple circuit structure, over the entirety of a necessary tonal range without being affected by variations of dark output values during the read-in of the actual picture.

The above and other objects, features and advantages of the present invention will become apparent from the following description and the appended claims, taken in conjunction with the accompanying drawings, in which: Figure 1 is a block diagram of a picture input equipment incorporating the present invention; Figure 2 illustrates a multi-elements photosensor equipped with light-shielded portions; Figure 3 is a diagram of a D.C. clamping circuit which serves to detect dark output voltages (i.e., voltages from the shielded portions of the multi-elements photosensor) and then to subtract the voltages from their corresponding picture signals; Figure 4(a) is a waveform diagram showing signals output from the multi-elements photosensor when a picture of a uniform density is read in;; Figure 4(b) is a waveform diagram of voltages output from a sample holding circuit after dark output voltages have been subtracted from the signals of Fig. 4(a); and Figure 4(c) is a waveform diagram of normalized output voltages obtained after correction where correcting reference signals and actual picture signals are identical to each other.

The present invention will hereinafter be described by the following specific example.

The photoelectric conversion characteristics of each element of a photosensor such as photodiode array, CCD (Charged Coupled Device) or the like satisfy, as well known in the art, the following relationship: Ví=S.Xj+ D; (1) where V output voltage of the photosensor.

S,: conversion efficiency.

Xj the quantity of incident light.

D j: dark output voltage.

Variations in conversion efficiency among individual elements are usually within 10% or so. However, it is known that variations in output voltage, including the shading (vignetting of oblique beams) by a lens system, reach 30%-50%.

On the other hand, the dark output voltage is about 1 % of Vi or so and its variations are within about 10% or so.

For the correction of the above-described variations in output voltage, a picture having a uniform reflectivity or transmission and serving as a reference for white level is first of all illuminated to read in signals.

At this time point, the output voltage V, of the ith element of the multi-elements photosensor varies depending the quantity X" of incident light and is expressed by the following equation: V,l = Sl . X,l + Dn (2) Representing the dark output voltage D,, in terms of its representative value D,d (for example, a voltage output from a partially light-shielded portion at one end of the multielements photosensor) and their difference AD,,, the following equation is derived:: D@=D@d+#D@@ (3) Since the variation, i.e., difference AD,, in dark output voltage among individual elements is very small compared with the output voltage V, of the photosensor and may thus be ignored (AD,, < < V,), the above equation (2) can be rewritten as follow: V"~S,. Xn + D,d (4) Subtracting here the representative value D,d of the dark output voltage from a reference picture signal read in the above-mentioned manner, a correcting reference signal V, to be stored is expressed as follow: V,=V,j D,d = Sj.X,^ (5) When reading in an actual picture next, the quantity Xi of the incident light can be considered as a quantity obtained by subjecting the quantity X, of standard incident light to modulation (modulation index: K3 by the real picture. Accordingly, the output voltage Vj of each of the elements of the multi-elements photosensor, said output voltage Vi being expressed by the equation (1), can be rewritten as follows: V1= 51. K1. Kl+Dd + AD ~Sj. Kj. X,^+ Dd (6) In the equation (6), K, means a modulation index of the jth element to the reflectivity by or transmission through the real picture and Dd denotes the representative value of dark output voltages upon read-in of the real picture.

Subtracting here the representative value Dd of dark output voltages from the output voltage Vg, the signal voltage V^, to be corrected can be expressed by the following equation: Vi'= V,-Dd = Si. K@. X@ (7) The voltage V, of the real picture signal is subjected to a correcting operation such as division or the like by the correcting reference signal, thereby deriving the following equation: Vv/v"t= S,. K,. X@@/S,. X, = Kj After the correction, the output signal is a real picture signal which is proportional only to the modulation index K,. Thus, the effects pertaining to the sensitivity, dark output, lighting and optical systems and the like of the multielements photosensor have all been corrected, thereby obtaining a real picture signal the level of which is governed solely by the density (reflectivity or transmittance) of the original.

One embodiment suitable for use in the practice of correction of picture signals in accordance with the method of this invention will be described with reference to the block diagram shown in Fig. 1.

In the block diagram of Fig 1, there are shown a multi-elements photosensor 1 for reading in a picture, a dividing circuit 2 for performing a correcting operation, a DC clamping circuit 3 for detecting an output value (the representative value of dark outputs) of a photosensor element shielded from light at one end of a multi-elements photosen sor (for example, a multi-elements photosen sor such as that depicted in Fig. 2) and for subtracting the output value from a picture signal, a sample holding circuit 4, an A/D (analog-to-digital) converting circuit 5 for en coding the analog value of the correcting reference signal to its corresponding digital value, a RAM (random access memory) 6 for storing the correcting reference signal, a D/A (digital-to-analog) converting circuit 7 for con verting the encoded correcting reference sig nal to an analog value, a timing control circuit 8 of the picture input equipment, an R/W (read/write) control circuit 9 for reading data out from the RAM 6 or writing data in the RAM 6, and a switch 10 for inputting the correcting reference signal to the read-in sto rage unit.

Fig. 2 illustrates the outline of the photosensor 1. As indicated by hatching, lightshielded portions 21 are provided at both ends. Outputs from these portions are used as dark outputs.

Fig. 3 shows an exemplary structure of the DC clamping circuit 3. When inputting a picture, the level of the signal from the lightshielded portion of the multi-elements photosensor 1 is retained in the form of a lightshielded portion detection pulse (timing: to) until the next light-shielded portion detection pulse is generated at the sample holding circuit 11. It is then fed as a representative value of dark outputs to the subtracting circuit 12, where it subtracted as the representative dark output value from a picture signal.

Prior to read-in of an actual picture, a picture which serves to provide a reference for white level is first of all read in from the multielements photosensor 1. The output voltage V, of the photosensor 1 has such a signal waveform as shown in Fig. 4(a).

By the DC clamping circuit 3 having such a circuit structure as shown in Fig. 3 for example, signal value output from the lightshielded portion of the photosensor which signal value is contained in signals output from the photosensor is detected at the timing of 4. This value is retained in the sample holding circuit 11. At the subtracting circuit 12, this value is used as the representative value D,d of dark outputs so as to subtract D,d from the output voltage V, of the photosensor 1.

The value left after the subtraction is fed to the sample holding circuit 4, where the waveform of the output voltage is retained in the form of its corresponding timing pulse of the timing pulses t,,t,,..., for signals from respective picture elements as shown in Fig.

4(b). In order to store the timing pulse as the correcting reference signal in the RAM 6 by way of the switch 10, the timing pulse is subjected to an A/D conversion by the A/D converting circuit 5.

If the correcting reference signal is encoded with a resolution of 8 bits in order to perform a correcting operation, it is possible to obtain only 6-7 bits as density graduation of the thus-corrected picture signal due to operation errors, downward shifting of significant figures, etc.

When an 8-bit accuracy (256 stages) is required as density graduation for corrected picture signals, it is thus necessary to encode beforehand the correcting reference signal with a resolution of about 10 bits (1024 stages) or so by an A/D converting circuit 5 in view of the imminent downward shifting of significant figures through the operation.

In the above-mentioned manner, the correcting reference signal is stored in advance in the RAM 6.

Upon read-in of the actual picture, the switch 10 is kept open. In synchronization of the read-in of the actual picture by the multielements photosensor 1 and at the same time, in synchronization with the holding signal of the sample holding circuit 4, correcting reference signals which have been stored in the RAM 6 are read out successively in accordance with read-out command signals output from the timing control circuit circuit 8 and adjusted in read-out timing by the R/W control circuit 9.

Each correcting reference signal is converted by the D/A converting circuit 7 into a correcting analog reference signal V,? and is then fed to the dividing circuit 2.

From the real picture signal V" the representative value Dd of dark outputs is subtracted at the DC clamping circuit 3. Then, the remaining value is supplied via the sample holding circuit 4 to the dividing circuit 2, where it is divided by the above-described correcting reference signal V",.

Fig. 4(c) shows the waveform of corrected output signals when signals identical to their corresponding reference signals are read in as actual picture signals. Variations among the individual elements of the photosensor, changes in dark outputs, illumination irregularity due to variations in the optical and lighting systems, etc. have been corrected, thereby resulting in a voltage of constant output.

Having now fully described the invention, it will be apparent to one of ordinary skill in the art that many changes and modifications can be made thereto without departing from the spirit or scope of the invention as set forth herein.

Claims (3)

1. A method for correcting picture signals in a picture input equipment adapted to read in an actual picture by means of a multielements photosensor, which method comprises: reading in a picture of a uniform density prior to the read-in of the real picture; encoding values which have respectively been obtained by subtracting dark output values of the individual elements of the photosensor from the values read above by the respective elements, and storing the thusencoded values as correcting reference signals for their corresponding elements; and upon reading in the real picture, reading out the correcting reference signals in synchronization with the read-in of the real picture so as to correct values, which have been obtained by subtracting the dark output values from the thus-read values of the real picture, on the basis of the above read-out correcting reference signals.

2. A method according to Claim 1, wherein signal values output respectively from light-shielded end portions of the individual elements of the multi-elements photosensor are used as the dark output values.

3. A method according to Claim 1, wherein when encoding the values, which have respectively been obtained by subtracting dark output values of the individual elements of the photosensor from the values read above by the respective elements, to obtain the correcting reference signals, the first-mentioned values are encoded with a resolution higher than a tone resolution required as an output.