JPH069374B2 - Image signal binarization circuit - Google Patents

Description

The present invention relates to a circuit for binarizing an image signal.

(Problems to be Solved by Conventional Techniques and Inventions) When an original image is represented by binary values of black and white, a binary method using a constant threshold value is used for images such as characters. ,
For multi-tone images such as photographs, the dither method using a plurality of threshold value sets is used.

When a multi-tone portion such as a photograph and a two-tone portion such as a character are mixed in the same document, if the dither method is used to express the multi-tone, the character becomes unreadable If this is set, there is a drawback that binarization of a multi-tone portion cannot be performed well.

Further, when the image is read, if the background density of the document changes, the image is also affected.

An object of the present invention is to provide an image signal binarization circuit capable of expressing an image in which a 2-gradation image and a multi-gradation image are mixed.

(Means for Solving Problems) An image signal binarizing circuit according to the present invention is a photoelectric conversion means for converting the density of an image of a document into an analog electric signal, and a digital image signal of a plurality of bits for the analog electric signal. An AD converting means for converting the digital image signal into a binary image by a binarizing process using a constant threshold value or a dithering process using a dither method; and a point to be binarized. Storage means for storing digital image signals of a plurality of surrounding n points, and an average for obtaining an average value of digital image signals of a plurality of m points (m ≧ n) surrounding the point including the point to be binarized Means and a digital image signal of the above n points, a difference between the digital image signal of the point to be binarized and the digital image signal of the other point is obtained, and the difference is within a predetermined value. The number of If there is the above and the difference between the digital image signal of the point to be binarized and the average value is equal to or more than a predetermined value, the binarization process using a constant threshold is adopted, and other than that. In this case, a binarization control means for averaging and adopting a binarization process by the dither method is provided.

(Operation) If there is no sharp change in the density near the binarizing point even within the same document, multi-gradation is expressed by the binarization processing by the dither method, and if there is a change, it is kept constant. Two gradations are expressed by the binarization process using the threshold value.

(Examples) Examples of the present invention will be described below with reference to the accompanying drawings. Here, FIG. 1 is a block diagram of an embodiment of the present invention, and FIGS. 2A to 2D are explanatory diagrams for explaining the operation.

In FIGS. 2A to 2D, reference numeral 31 is a platen glass on which an original 41 is placed, and white is applied to the end of the platen glass 31 to form a reference white 32.

In the optical system, the reference white 32 or the image of the original document illuminated by an exposure source (not shown) passes through reflecting mirrors 25, 24, 23, and a photoelectric conversion device (one-dimensional CCD) 21 by a lens 22.
The image is formed on the light receiving surface of the.

The reflecting mirrors 23, 24, 25 can move in the direction of arrow A. At that time, the reflecting mirrors 23 and 24 move at half the speed of the reflecting mirror 25, and the optical path length l is constant regardless of the position of the reflecting mirror 25.

When the reflecting mirror 25 is located at the position shown in FIG. 2 (a), the density of the reference white 22 is detected. That is, in the circuit of FIG. 1, first, the reference white level detection circuit 11 outputs a constant voltage, which is the reference voltage for AD conversion, to the reference voltage output 11a. The AD converter 1 uses this voltage as a reference voltage to convert the analog signal of the reference white density analog output 21a detected by the photoelectric converter 21 into an 8-bit digital signal, and outputs it to the digital output 1a. Then, the reference white level detection circuit 11 stores the value for one line from the digital output 1a, and ends the detection of the reference level.

Next, the reflecting mirrors 23, 24 and 25 are shown by arrow A'in FIG. 2 (b).
The original 41 is sequentially scanned. At this time,
The reference white level detection circuit 11 outputs the stored reference white level to the reference voltage output 11a. Therefore, it is possible to correct variations in each element of the photoelectric converter 21 and shading, and then convert the signals into digital signals by the AD converter 1. Further, even a document having a high background density can be effectively binarized. The peak value detection circuit 12 detects and stores the peak value in a predetermined area of the document 41.

The reflecting mirror 25 is moved and the end of the original 41 is shown in FIG. 2 (c).
When the position is reached, the scanning system stops.

Therefore, the peak value detection circuit 12 outputs the peak value output 12a.
The reference white level detection circuit 11 corrects the stored reference white level based on the input peak value and outputs the corrected reference white level to the reference voltage output 11
Determined as the value of a.

Next, as shown in FIG. 2 (d), the reflecting mirror 25 moves in the direction of arrow A ″ to read the original. At this time, the digital output 1a is corrected by the above-mentioned peak value and reference white level. The 8-bit digital value is output, and the average value circuit 2 calculates the average value from the digital value and the necessary value read from the memory 3 and outputs the average value to the average value output 2a. The digital value output from the AD converter 1 is written.

This average value is an average value of an 8 × 8 matrix (matrix) including points to be binarized. Therefore, the memory 3 has a capacity of storing at least eight lines including a line having a binarization point, and is configured to update and store the subsequent lines in sequence as the binarization of each line is completed. Has been done. The average value circuit 2 reads the data in the memory 3 and calculates the average value.

From the memory 3, the value of the previous line of the line including the binarized point is output to the memory output 3a, and the value of the same line is output to the memory output 3a.
In b, the value of the next line is output to the memory output 3c. Then, the shift register 4 stores the values of the points to be binarized in the shift register output 4e, and the values of the points around it are stored in the shift register outputs 4a to 4d and 4f to 4i.

A logic circuit 5 including a microprocessor, ROM, and RAM performs binarization from the average value output 2a from the average value circuit 2 and the shift register outputs 4a to 4i. That is, the logic circuit 5 obtains the difference between the shift register output 4e and the other shift register outputs 4a to 4d and 4f to 4i, and if the difference is within a predetermined amount, the number of the above differences is two or more. Judge as not noise. Then, when the shift register output 4e is not noise and the difference between the shift register output 4e and the average value output 2a is a predetermined amount or more, the value of the binary output 5a is binarized using a constant threshold value. It is set to 0 or 1 by processing. In other cases, the value of the binarized output 5a is determined by the dither method according to the value of the average value output 2a and the position of the point to be binarized.

In this embodiment, an 8 × 8 matrix including binarization points is used in order to simplify the circuit for calculating the average value.
The size of the matrix may be determined according to the image or the like. Also, the number of differences may be appropriately determined according to the image or the like.

(Effect of the invention) When there is no abrupt change in density near the binarizing point, gradation is expressed by binarization processing by the dither method, and when there is a change, gradation is expressed. As a result, characters and the like can be clearly binarized in two gradations.

Further, since noise on the document or noise on the circuit is removed, an image without noise can be obtained.

Since the processing is performed after conversion into a digital signal, an inexpensive digital memory can be used.

[Brief description of drawings]

FIG. 1 is a circuit diagram of an image signal binarization circuit. Fig. 2
(a)-(d) is a figure which shows the position of the optical system in image reading one by one. 1 ... AD converter, 2 ... average value circuit, 3 ... memory, 4, 4, 4 ... shift register, 5 ... logic circuit, 21 ... photoelectric converter, 31 ... platen glass, 32 ... ... Standard white, 41 ... Original.

Claims (1)

[Claims] 1. A photoelectric conversion means for converting the density of an image of an original into an analog electric signal, an AD conversion means for converting the analog electric signal into a digital image signal of a plurality of bits, and a binary value using a constant threshold value. Binarizing means for binarizing a digital image signal by binarizing processing or binarizing processing by the dither method, and storage means for storing digital image signals of a plurality of n points surrounding the point to be binarized. And a plurality of m points around the point including the point to be binarized (m ≧
n) averaging means for obtaining an average value of the digital image signal, and for the n-point digital image signal, the difference between the digital image signal at the point to be binarized and the digital image signal at the other points is obtained. , The difference is within a predetermined value, the number of differences is a predetermined value or more, and the difference between the digital image signal at the point to be binarized and the average value is a predetermined value or more. Use a constant threshold for 2
An image signal binarization circuit, comprising: a binarization control unit that employs a binarization process and, in other cases, averages the binarization process by a dither method.