JPH02214264A - Picture processor - Google Patents

Abstract

PURPOSE:To prevent leakage of secrecy by providing a means modifying an input picture into a picture illegible by visual observation and outputting the resulting picture and a means reading the modified picture and decoding it into the same picture as the input picture. CONSTITUTION:A selector 257 select an input terminal of a signal outputted to a terminal D from any of A, B and C depending on a signal inputted to a terminal E. In the case of picture modification copy, the terminal A is selected and the signal is outputted to the terminal D, then a data of a random number ROM 255 is outputted, different values are obtained every time signals B, D enter. In the case of picture decoding copy, the terminal B is selected and the signal is outputted to the terminal D, then the data from the random number ROM 255 is outputted and different values are obtained every time signals B, D enter. Thus, the write position for each line is revised according to the random number to generate a illegible picture or a normal picture is decoded from the illegible picture. Then the leakage of secrecy is prevented.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to an image processing device that reproduces an input image by deforming it so that the original image cannot be imagined from the reproduced image.

[Conventional technology]

Conventionally, copying machines have been designed to reproduce images that are faithful to originals.

Various types of copying machines have also been manufactured that can transform images, such as enlarging or reducing the image, and independently varying the enlargement or reduction ratio in the vertical and horizontal directions.

[Problem to be solved by the invention] However, since the degree of deformation in these image deformations is relatively small, it is possible to imagine the original image from the deformed reproduced image, and even character images. does not interfere in any way with reading the meaning of the text.

Therefore, it was powerless for the purpose of not wanting image information to be known to anyone other than specific people.

[Means for solving problems]

According to the present invention, by changing the writing start position for each line according to random numbers, it is possible to create an illegible image or to restore a normal image from an illegible image.

〔Example〕

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

FIG. 1 shows an example of a schematic internal configuration of an image reading device according to the present invention. Further, FIG. 2 is a timing chart showing the operation of each part of the circuit of FIG. 1.

The original placed on the platen glass is exposed and scanned by a mirror unit consisting of a halogen lamp and a reflecting mirror, and the reflected light image from the original is scanned by a lens unit.
A reduced image is formed on the line color CCD sensor 103.

The color separation image signals read every one line without exposure scanning are inputted to the video processing circuit shown in FIG. 1 and subjected to signal processing. A signal 214 is a signal line for driving the 3-line color CCD sensor 103, and necessary driving pulses are generated by the pulse generator 201. The video signal processing circuit shown in FIG. 1 is controlled by a CPU (not shown) via a bus 215.

Next, the video processing circuit will be described in detail. The original is first irradiated by an exposure lamp, and the reflected light passes through a mirror unit and an optical lens to form a reduced image on the 3-line color CCD sensor 103, which is separated into colors for each sensor and read, and sent to the amplifier circuit 202. 203 and 204 are amplified to a predetermined level. The color CCD sensor used in this example has 3 lines of line sensors arranged in parallel at a distance of 180 μm (18 lines), each having 10 μm x 10 μm and 5000 effective pixels, each with R2G, R2G, It is dyed with organic dye B. Further, each COD is driven independently and synchronously. Video signals VR, Vo, V8 are sent from each CCD 103a, 103b, 103c in synchronization with the drive pulse.
are independently output, and independent amplifiers 202.203.20
After being amplified to a predetermined voltage value in step 4, the A/D conversion circuit 20
5. 206 and 207, each is converted into 8-bit digital data.

Next, in this embodiment, as described above, each CCD has an interval of 18 lines (10 μm x 18 = 180 p m) in the sub-scanning direction, and when viewed from the CCD 103a of the red (R) filter that scans in advance, it is green (G). ) Filter C
CD103b and CCD103 with blue (B) filter
The spacing between CCDs 103a and 103 is 18 lines (180 μm) and 36 lines (360 μm), respectively.
b.

The reading position of 103c is shifted. Therefore, in order to connect this correctly, we use memory for multiple lines.

In other words, in order to synchronize with the B-CCD 103a that is read last, if the sub-scanning reading density is 400 dpi, in the case of 100% same-size reading, the R-COD buffer memory 208 has 36 lines worth of space, and the G-CCD buffer memory 209 has a synchronization memory for 18 lines. Furthermore, considering the maximum enlargement scan of 400%, the number of lines in each buffer memory 208 and 209 is 3.
6×4=144 lines, 18×4-72 lines are required. Here, the number of pixels for one line is the effective number of pixels, which is a minimum of 4678 pixels, assuming that an A4 size sheet of 297 mm in the longitudinal direction is read at 400 dpi. Then, the last read B-CCD 103c reads the R-CCD 1 first.
03a, the buffer memories 208 and 209 synchronize with the reading of the same line as the line read by the G-CCD 103b.
By reading data of each color on the same line, video data in which the same line is separated into three colors (R, G, B) is obtained.

Next, the black level correction circuits 210, 211, and 212 will be explained.

Prior to the copy operation, the black level image signal is adjusted to the black level correction circuit 210°211.21 without turning on the halogen lamp.
2 and correct the bit-by-bit variation of the line sensor at this time so that they become the same value. This is to ensure that all outputs for one line have the same value when reading a low-density portion of a uniform image.

Next, white level correction (shading correction) is performed on the signal subjected to the black level correction by the white level correction circuit 213.
It will be held at 214.215. White level correction is based on white data when a uniform white plate is illuminated, and the lighting system, optical system,
Correct sensor sensitivity variations, etc.

As described above, the black level and white level are corrected based on a number of factors such as black level sensitivity of the image input system, dark current variation, optical system light intensity variation, white level sensitivity, etc., and the black level and white level are made uniform across the main scanning direction. Color image data proportional to the amount of input light is input to logarithmic conversion circuits 216, 217, and 218 in accordance with the relative luminous efficiency characteristics of the human eye.

Here, the data output for each B, G, and H corresponds to the density value of the output image, and for B, the amount of yellow (Ye) toner, and for G, the amount of magenta (M). The toner amount corresponds to the cyan (Cy) toner amount for R, so from now on, the color image data will be Ye, M,
Map it to Cy.

Color correction 2 for each color component image data from the original obtained by logarithmic conversion, that is, Ye acid component, M component, and cy component.
19 will be held. Color〇The spectral characteristics of the color separation filter placed in the CD sensor have an unnecessary transmission region, and the color toner (Y, M, C) transferred to the transfer paper also has unnecessary absorption components, so each color component Image data Y+, M5
A masking circuit 219 performs masking correction that calculates a linear equation for each color (C and white) and performs color correction.

Furthermore, Min (Y+,
M+, C+) (minimum value of Y+, M+, C+) is calculated, this is used as a black mark, and then black toner is added (smear insertion) and each color material is adjusted according to the added black component. A black extraction/UCR circuit 220 performs an under color removal (UCR) operation to reduce the amount added.

Each color signal that has been subjected to masking correction, UCR, and shading processing is subjected to γ correction processing in a γ correction circuit 221, and if necessary, scaling control in the main scanning direction is performed in a scaling processing circuit 222, and then filtered. A sharpness circuit 223 performs edge enhancement and smoothing processing. These processes are performed by digital arithmetic processing.

Next, the 8-bit image signal that has been subjected to the various processes described above is converted into 1-bit data by a binarization circuit 224 that performs, for example, dither processing.
It is converted into a signal of either 0 or 0.

Next, a circuit for transforming an image will be explained in detail.

RESET signal 252 (
F/F252 sets the output terminal to L (Fig. 2 ■) (
Figure 2 ■). As a result, line memory A250 is
In state D (■ in Figure 2), line memory B251 is in WRI state.
The state becomes TE (Fig. 2 ■). On the other hand, the polygon mirror 227 is always rotating at a constant speed, and the laser driver 225
The laser light emitted from the semiconductor laser 226 is
The photosensitive drum 228 and the light receiving element 229 are scanned. When the light receiving element 229 is irradiated with the laser, it receives B and D signals (second
■) in the figure is sent to the synchronization circuit 254.

The synchronous circuit 254 converts the B and D signals to VCLK (Fig. 2 ■)
PH5YNC signal generated in synchronization with (Fig. 2■
) as address counter 253 and preset counter 2
Send to 58.

This PH3YNC signal causes the address counter 253 to count up (FIG. 2). The output of the address counter 253 is input to an address terminal of a memory in which random numbers are written in advance, for example, a random number ROM 255. The random number ROM 255 outputs the stored data at the designated address (Fig. 2, ■).

This random number data is input straight to the A terminal of the selector 257, and inverted by the inverting circuit 256 to the B terminal. Furthermore, 01H (hexadecimal number) is always input to the C terminal by pull-up and pull-down.

The selector 257 is the signal input to the E terminal (this is the CP
The input terminal for the signal to be output to the D terminal is selected from A, B, or C using the switch (which may be from U or may be a manual switch).

First, A is an image transformation copy mode.

In the case of image modification copying, select the A terminal and output to the D terminal. Since the data from the random number ROM 255 is output, the value will be different each time the B and D signals are input. For example, if the address and data of the ROM 255 are as shown in Table 1, data 51H at address 0OOOH is input to the D terminal of the preset counter 258 until the first PH5YNC is input, and the first B and D are input to the D terminal of the preset counter 258. The moment 51H is input to the terminal, it is preset (Fig. 2 @).

Next, each time VCLK is input to the CLK terminal, it counts down, and when VCLK is input 51H times, the preset counter becomes O (Fig. 2), and a borrow signal (Fig. 2 @) is output from the BO.

Similarly, when the second shot B, D (0 in Figure 2) comes in,
After OHVCLK is input, a borrow signal (0 in Figure 2) is output from BO.

On the other hand, the binary image signal sent from the binary circuit 244 in synchronization with VCLK is first sent to the line memory B251.
They are written in the order they were sent. Then, the moment the BO ratio output of the preset counter 258 is input to W-RES, the mode switches to READ mode (0 in Figure 2), and the written contents are output from the OUT terminal in order from address OO. Yes]) Laser driver 255
send a signal to The laser driver controls the semiconductor laser 226 in 0N10FF based on the signal and emits laser light. The laser beam is deflected by a polygon mirror 227 and scans over a photosensitive drum 228.

In this way, a cyan image is created from the image signal read by the R-CCD 103a in FIG. 1, a magenta image is created in the same way, and a yellow image is created to create a full color image.

The reason why there are two line memories A250 and B251 is to alternately switch between writing to one and reading from the other every time the B and D signals are received.

Next, in the case of image restoration copying, the B terminal is selected and the data is output to the D terminal. Since data from the random number ROM 255 is output, the values will be different each time the B and D signals are input.

For example, if the address and data of the ROM 255 are as shown in Table 1, the value AEH obtained by inverting the data 51H at address 0OOOH is input to the D terminal of the preset counter 258 until the first B and D are input. Then, AEH is preset the moment it is input to the P terminal. Next, each time VCLK is input to the CLK terminal, it counts down, and when VCLK is input AEH times, the preset counter becomes O and a borrow signal is output from BO.

Similarly, when the second B and D are input, a borrow signal is output from BO after the value 5FHVCLK, which is the inversion of AOH, is input.

For normal copying, select the C terminal and send 01H to the D terminal.
(hexadecimal) is output. This is preset in the preset counter 258 in synchronization with the B and D signals, and then the C
When VCLK is input once to the LK terminal, the counter value becomes 0.
becomes 0 and generates a borrow signal. In other words, in normal copying, the borrow signal lags behind the B and D signals by only one clock.

As explained above, in the case of a random number ROM as shown in Table 1, the writing start position on the photosensitive drum based on the B and D signals in deformation mode and restoration mode is as shown in Table 2. .

When a document is copied, the lth line is copied with a delay of 51H in the modified copy mode.

If the copy is copied in restoration copy mode, it will be copied with an AEH delay.

In other words, the original is copied with a delay of 51H+AEH=FFH. Similarly, the second line is AOH+
5FH=FFH 13th line is 9DH+62H=FFH
Each is copied with a delay.

In this way, in the modified copy mode, the start of writing for each line is 51H, AOH in terms of the number of clocks.

9DH,..., the copied image will be completely different from the original because it will be scattered, but if you copy this in restoration copy mode, the start of writing for each line will be delayed by FFH in clocks. In order to align the images, the same image as the original is copied with an overall shift of FFH. This results in an image that can be easily understood by humans.

[Other Examples]

By the way, a similar effect can be obtained by switching between outputting the image signal as it is or inverting it for each minimum pixel unit using random numbers.

Figure 3 shows its configuration. It should be noted that blocks similar to those in FIG. 1 are given the same numbers and their explanations will be omitted.

In other words, if the screen of the original is A3, it is 420mm x 297mm.
mm, when creating an image from 16 dots per mm, 4
20mmX16X297mmX16=31933440
Therefore, 31933440 dots are required. This is 1
If converted into a hexadecimal number, it becomes IE74400, which can be covered with 25 BIT, so the address counter 501 is 25 BIT.
shall be.

This address counter 501 is set to RESET for each color.
The output is set to 0OOOOOOOH by signal 502 and counted up by VCLK. Random number RO
M503 is a ROM in which 1-bit random numbers are stored in advance as shown in Table 3, for example.

When the output of the random number ROM 503 is L, the image signal output from the binarization circuit passes through the no-inverter 505 and is sent as is to the laser driver 508.

When the output of the random number ROM 503 is H, the inverter 504
The output becomes L, and the image signal output from the binarization circuit when passing through the inverter 506 is inverted and sent to the laser driver 508.

In other words, when the output of the random number ROM 503 is H, each pixel of the image copied to the original is reproduced as black and white, but when the output of the random number ROM 503 is L, One pixel of an image that is copied is reproduced by inverting black to white and white to black.

The above-described image transformation and restoration can also be achieved by storing all image signals from a document once read into the memory of an external computer, as shown in Figure 4, and changing the color by software calculation according to random numbers in each small area. .

For example, the random number 00 is unchanged, and the random number 01 is yellow (Y
) to magenta (M), M to cyan (C), and C to Y. And at the time of 02, Y is CSC
It is also possible to convert M into M and M into Y, respectively.

Table 1 Contents of Random Number ROM [Effects of the Invention] As explained above, according to the present invention, it is possible to obtain an illegible copy and to restore a legible copy from the copy. Became.

This makes it possible to copy confidential documents and the like in an unreadable state and store them, thereby preventing leakage of secrets, and also making it possible to arbitrarily obtain the original image as needed.

[Brief explanation of the drawing]

FIG. 1 is a block diagram of an image processing device to which the present invention is applied;
2 is a timing chart of the circuit shown in FIG. 1, FIG. 3 is a block diagram of another embodiment, and FIG. 4 is a block diagram of still another embodiment, in which 224 is a binarization circuit, 225 is a laser 226 is a semiconductor laser 227 is a polygon mirror, 228 is a photosensitive drum, 229 is a B, D detection element, 250 is a line memory A, 251 is a line memory B, 252 is an F/F (flip-flop), 253
is an address counter, 254 is a synchronization circuit, 255 is a random number ROM, 256 is an inversion circuit, 257 is a selector, 258
is a preset counter.

Claims (7)

[Claims] (1) A means for deforming an input image so as to make it visually unreadable according to a predetermined deformation instruction signal and outputting the image; and a means for reading the deformed image and restoring it to the same image as the input image. An image processing device comprising: (2) Claim (1) is characterized in that the transformation instruction is a random number stored in a memory in advance, and the image output start position is changed line by line and output using this. image processing device. (3) Claim (1) is characterized in that the restoration instruction is a random number stored in a memory in advance, and the image output start position is changed line by line and output accordingly. image processing device. (4) The image processing apparatus according to claim (1), wherein the transformation instruction and the restoration instruction are stored in a memory having the same content. (5) The image processing apparatus according to claim (1), wherein the transformation instruction is a random number stored in a memory in advance, and the image signal is inverted based on the result. (6) The image processing apparatus according to claim (1), wherein the restoration instruction is a random number stored in a memory in advance, and the image signal is inverted based on the result. (7) The image processing device according to claim (1), wherein the transformation and restoration are color change and color restoration.