JPS59207782A - Shading correcting system - Google Patents

Abstract

PURPOSE:To make shading correction stable and accurate so as to improve the speed of facsimile communication with a picture reader by reading out shading information from a memory which is exclusively used for thisn purpose. CONSTITUTION:The image of a picture on an original is formed in a solid-state image sensor 6 by means of an optical lens 5. The image is photoelectrically transduced and a picture signal which received shading distortion is outputted. The picture signal is shifted at a clamping circuit 12 to adjust its DC level and inputted into a digital attenuator 13. At the attenuator 13, the inputted picture signal is multiplied by correction data read out from a memory EPROM14 and a corrected picture signal is outputted.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to an edding correction method for effectively correcting shading distortion in image reading devices, particularly facsimile devices.

In general, facsimile machines that read documents using an optical system that includes a solid-state image sensor (hereinafter referred to as CCD), such as a CCD image sensor, suffer from a decrease in the amount of light at the periphery of the lens (the so-called cosine fourth power law) and the amount of light from the fluorescent lamp that serves as the light source. Due to non-uniform distribution (ie, illuminance on the original), the photoelectrically converted image signal lacks uniformity as shown in FIG. The figure is a simulated representation of the image signal after photoelectric conversion. Figure 1 shows the image signal when an all-white area is read, but in the actual read signal, the arrangement of black and white areas on the document is irregular, and the period of white level (or black level) is extremely narrow. In some cases, the level between adjacent pulses may not drop to the level shown in Figure 1.
If the threshold value for black and white determination is constant, the threshold value cannot be set to a value close to GND, and both sides of the document are determined to be black. Therefore, in order to faithfully convert a monochrome original into a 2-short signal, the white level must be flat at any position on the original as shown in FIG.

However, as facsimile communication becomes faster, even if a fluorescent light is turned on when an original is set, scanning of the original must begin several seconds later. As a result, a transient state in which the light emission of the fluorescent lamp is not sufficiently stable, that is, a period t1+t2+...j7 r j8 where the light intensity distribution in the main scanning direction changes over time as shown in FIG.
The image must be taken with

Conventional shading correction of this type includes the following methods. The first method is a mechanical shading correction method, as shown in FIG. 4, in which a shading plate 4 is placed in the middle of the optical path to reduce the amount of light in the central portion to make the light amount distribution uniform. However, optical lens 5, fluorescent lamp 2, CCD 6
There are variations among facsimile reading devices including the above, and the adjustment of the shading plate 4 must be performed for each device. Adjusting the shading plate 4 requires processing such as cutting, which is a problem in terms of man-hours, and since the amount of light intensity correction is fixed, it is difficult to adjust the shading plate 4 due to the temporal changes in the light intensity distribution of fluorescent lamps as described above. The drawback is that it cannot be handled at all. In other words, if only one grading correction is applied when the light changes over time as shown in Fig. 3, the light amount distribution at each elapsed time 11-18 changes as shown in Fig. 5, and the shading correction is applied. The effect is seen only after time t8 when the fluorescent lamp becomes stable. This is because the center of the fluorescent lamp is cold, and while the electrodes at both ends heat the lamp and stabilize it quickly, the temperature rises slowly in the center, and it takes time for the light intensity to stabilize.

The second method is to read the reference white part immediately before imaging the original, create shading correction data based on the CCD output, and create shading correction data in random access memory (hereinafter referred to as R).
AM), and during the subsequent document imaging period, R
There is a method of performing shading correction based on data stored in AM.

According to this method, adjustment for each device is no longer necessary, and the drawbacks of the first method are solved. However, when transmitting multiple manuscripts in succession as shown in Fig. 6, it is necessary to leave an interval between the previous page manuscript 10 and the next page manuscript 11 in order to read the reference white part 9. However, this interval becomes an invalid time during communication, which is not preferable. Unfortunately, a facsimile machine with an automatic paper feed function has a complicated structure because it must use a clutch mechanism to control the spacing. Furthermore, the temporal change in the light intensity distribution as shown in Figure 3 occurs even during the transmission of the first page of the original, so the shading correction data stored in 1'LAM before the start of imaging cannot be expected to be effective. It becomes something.

Furthermore, as a third method, there is a method in which the white level is detected from the image signal being captured, and shading correction is performed based on the white level using the second method described above.

According to this method, it is possible to cope with changes in the light amount distribution during document imaging, and the drawbacks of the second method are solved. However, a white part is not always included in the image signal, and the detection of the white part is generally complicated and has the drawback of increasing the circuit scale.

The present invention provides a shading compensation method that can solve these drawbacks by performing semi-fixed shading correction.

The present invention will be explained in detail below.

FIG. 7 is a block diagram of an embodiment of the present invention.

The image on the document is captured by the optical lens 5! 1) An image is formed on the CCD 6, which photoelectrically converts it and outputs an image signal subjected to various shading distortions as shown in FIG. Since this image signal has a DC component superimposed on it, it is shifted by a clamp circuit 12 in order to adjust the DC level with a digital attenuator (hereinafter referred to as DATT) 13 at the next stage. DATT
The image signal input to 13 is stored in an EPROM (Etect
ricaffy ProgrammaMe ReadO
The correction data read from the DATT13 is multiplied by the correction data read from the DATT13.

The DATT 13 is composed of an 8-bit D/A converter as shown in Fig. 8, and the output of the clamp circuit 12 is input to the reference voltage input terminal VR of the D/A converter, while the 8-bit digital input terminal BITE is input to the reference voltage input terminal VR of the D/A converter. ~ BIT8 is an analog output terminal ■ by inputting shading correction data. A more corrected image signal is obtained. The shading correction data is stored in the EPROM 14, and each data is addressed by an address counter 15. Further, the address counter 15 counts (increments) the clock pulses synchronized with the CCD drive clock output from the timing clock generation circuit 17.

However, in this embodiment, the EPROM 14 has a plurality of seven
The lighting correction curves are stored in memory, and each correction curve is selected sequentially according to the elapsed time after lighting. It is measured in a control circuit 18 that can control. The control circuit 18 operates for a predetermined time after lighting, that is, in this embodiment, as shown in FIGS. 3 and 4, tl...t8
One pulse is output to the address counter 16 each time lapses. The counting state of the address counter 16 changes (increments) each time a pulse is input, and one of the plurality of shading correction curves stored in the EPROM 14 is selected corresponding to the counting state. Also, when the power is turned off or the communication ends and the fluorescent lamp 2 goes out, the control circuit 18 outputs a reset signal to the address counter 16, and when the fluorescent lamp 2 turns on again, 1. The data of the shading correction curve corresponding to is selected.

The operation of the control circuit 18 described above is shown in the flowchart of FIG. In the figure, Sl is 1. Similarly, 821S31S4185 is the correction curve used for the time interval from to t2.
．． 86. S7, and the correction curve used for the time after t8 is 88. tllj21... To judge each elapsed time of t8, the timing generation circuit 17 in FIG. 7 receives the black signal whose time in one line of the image signal is one cycle, and counts it. This has been realized due to certain things.

Next, the contents of the EPROM 14 will be explained. In this embodiment, in order to save the memory capacity of the EPROM 14, one correction data is given with four pixels as one unit, and strictly speaking, the white level of each pixel is not uniform.
This has little effect on the image signal after correction. Furthermore, 8 bits was used as the resolution in the amplitude direction for the tuning correction, as this value allows quantization errors to be sufficiently ignored. Therefore, since the correction data consists of 8 bits, the memory capacity required to obtain one shading correction curve is 1728 pixels, assuming that the number of main scanning pixels is 1728 pixels.
X-=432 (bytes). To specify this many addresses, 9 address lines (512 bytes) are required, which results in a remaining 80 bytes (512 bytes).
-432=80) requires a complicated address selection method, so it is not used to simplify the circuit configuration. In this embodiment, since a 4 kilobyte EFROM is used, there are 12 address lines. Therefore, the remaining three address lines allow selection of eight different seven-dimensional correction curves.

The address structure inside the EPROM will be further explained using FIG. 9. Each correction data D1...D432 is B
1...B8 consists of 8 bits, and each correction curve Sl...B8 is DI...... as mentioned above.
- Consists of 432 bytes of correction data of D432. In addition, the correction curve Sl...B8 corresponds to the elapsed time tbi...t8 after the lighting shown in FIG. 5, and the time t8 when the fluorescent lamp becomes sufficiently stable. Thereafter, the correction is continued using the data of the correction curve S8. EPROM
There are several possible methods for writing correction data into the facsimile machine, but in this embodiment, this is done by preparing a dedicated correction data creation device outside the facsimile machine. The correction data creation device inputs the photoelectric conversion output of the facsimile device to which shading correction is to be performed, and performs the steps 1 and 1 in FIG.
- measure the 7 aging distortion at each time of 18;
Create correction data for each.

After that, this correction data is transferred to EPRO using a ROM writer.
Write in M. The written EPROM is inserted into a socket in the aging correction circuit of the facsimile machine.

This ROM writer can take any form and can be used.

As explained above, according to the present invention, since the shading correction information is read from the EPROM without being extracted from the image signal, stable and accurate shading correction can be easily performed, and the shading correction information can be easily performed over time. Appropriate corrections are made, so it can also support faster Facun Milli communications.

[Brief explanation of drawings]

Figure 1 is a time chart that simulates an image signal without shading correction, Figure 2 is a time chart that simulates an image signal that has undergone shading correction, and Figure 3 is a time chart of light intensity distribution. Figure 4 is a system diagram showing the configuration of a conventional mechanical shading correction method, and Figure 5 is a time chart showing temporal changes in image signals when performing conventional mechanical shading correction. , 6th
The figure is a system diagram showing an example of a configuration for reading and correcting white areas using a conventional shape ink correction method, FIG. 7 is a block diagram of an embodiment of the present invention, and FIG.
FIG. 9 is a schematic diagram showing an address structure of an EPROM used in the present invention, and FIG. 10 is an operation flowchart showing the operation of a control circuit used in the present invention. 11- Shoulder 1 Figure porridge 2 Figure 3 Figure □GND porridge 8 Seki-491- Tsumugi 10 Flash

Claims (1)

[Claims] A plurality of pieces of shading correction curve data corresponding to changes in illuminance on the document over time after the light source is turned on are stored in an erasable and writable read-only memory, and the data are stored in the read-only memory according to the time elapsed after the light source is turned on. Shading characterized by specifying an address, sequentially selecting and reading out the plurality of corresponding shading correction curve data, and performing shading correction by multiplying the photoelectrically converted image signal and the selected shading correction curve data. Correction method.