JPS605383A - Image read system - Google Patents

Abstract

PURPOSE:To make a precision read through manual feeding operation by providing two arrays of linear image sensors, and thinning out scanning lines according to the rate of the moving speed and basic moving speed of the sensors make correction. CONSTITUTION:The linear image sensors 9-1-9-2 arrayed at an interval l are moved as shown by an arrow 4 to read a character 3 written on an original 2. An up-to-down scan is made repeatedly in the sensors 9-1-9-2 to make a read repeatedly. The ratio N/N0 of the number N of scans which are made until the sensor 9-2 reaches the same position with the sensor 9-1 and the basic number N0 of scans expected in manual feeding operation is calculated. Then, thinning-out information is obtained from the calculated data and supplied to a one- character memory circuit stored successively with the photoelectric transducing output of the sensor 9-1 to thin out scanning lines at intervals of a specific number of lines, thereby generating an output corrected to the correct size as a character form.

Description

DETAILED DESCRIPTION OF THE INVENTION (1) Description of the field to which the invention pertains The present invention detects characters, symbols, etc. written on a document by hand-feeding a reading device while keeping the document fixed, and detecting photoelectric conversion signals. The present invention relates to a simple image reading method that can obtain images.

(2) Description of Prior Art FIG. 1 shows a reading device that performs a manual feed operation. In the reading method for manual feeding operation, the operator holds the reading device 1 in his hand and moves it along the document 2 in the direction shown by the arrow 4, for example, to read the characters 3 on the document, and the reading signal is transmitted through the signal line. 5 leads to recognition of Oribe. This reading method has already been applied to OCR and the like.

A diagram of the principle of this type of conventional reading method is shown in FIG. In the figure, 6 is a two-dimensional image sensor, 7 is a lens optical system, and 8 is a light source lamp.

The two-dimensional image sensor 6, which is shown in a mosaic pattern in Figure 9, is usually composed of 40 x 12 bits (vertically and horizontally), and by adjusting the magnification of the lens optical system 7, one character (for example, 0CR-B (character size of the font) is stored in the two-dimensional image sensor. The light source lamp 8 is for illuminating the surface of the document.

3(tt), (b), and (C)) show the reading state in terms of the positional relationship between the letter "A" and the sensor 6.

tOs tl+t2 represents each time. The two-dimensional image sensor 6 moves at time t according to the moving direction of the arrow 4. -+1. → t2, and when the state at time t (=) is reached, if the output of the image sensor is taken out, the complete character *A n can be read.

This state is detected when there are spaces on both sides of the character A'', and this state indicates that one character is located at the center of the image sensor.

This can therefore be achieved independently of the operator's feed rate. In other words, in the case of manual feed operation, the manual feed speed differs depending on the operator.

It is necessary to be able to read reliably at any speed.

However, this method has the disadvantage that the reading device cannot be constructed at low cost because the two-dimensional image sensor is extremely expensive.

(3) Purpose of the Invention The present invention solves these drawbacks by providing and combining two lines of inexpensive one-dimensional image sensors (achieving a highly accurate reading system using two-to-nine hand-feeding operations). will be described in detail below with reference to the drawings.

(4) Explanation of structure and operation of the invention FIG. 4 shows an embodiment of the present invention, and shows a schematic diagram of a reading method. Reference numbers 2 through 8 are used in common with those shown in FIG. Also 9-1, 9-
2 are one-dimensional image sensors, ctL
Thin film photoelectric conversion elements such as z, Se, etc., or COD, MO
An 8i element such as S is used. The lens optical system 7 is a light guide element such as a normal spherical lens or a convergent fiber array.

The light source lamp 8 is a tungsten lamp, fluorescent lamp or LE.
D solid-state light sources are used. By moving the one-dimensional image sensor 9-1.9-2 in the moving direction shown by the arrow 4, the one-dimensional image sensor 9-1.9-2 passes through the lens optical system 7 and reads the characters written on the original 2. Read 3. Further, the interval 1 between the one-dimensional image sensors 9-1 and 9-2 is determined based on the size of the characters written on the document and the magnification of the lens.

FIG. 5 shows a conceptual diagram for explaining the reading operation.

interval! One-dimensional image sensor 9-1.9- arranged in
2 is assumed to be moving in the direction of arrow 4. Also 1
In the dimensional image sensors 9-1 and 9-2, two images are scanned, for example, from the top to the bottom of the figure.

Reading is being performed repeatedly.

In the moving state (α), when sensor 9-1 detects 1 black, sensor 9-2 detects the same 6 black as detected by sensor 9-1 in moving state (α). ” is detected. Basic speed ■ in the direction of arrow 4. The basic number of scans N when changing from so-called state (α) to ()) until sensor 9-2 reaches the same position as sensor 9-1 when moved. is given by N6=nX1. However, le is the linear density in the direction of movement. Next, let V be the moving speed by manual feed operation.

Let N be the number of scans until the sensor 9-2 reaches the same position as the sensor 9-1, and let τ be the scan period.

! =V, N, τ=V. ＃　No＊τ Because it holds true.

V = No/N, V.

It can be expressed as From this, the ratio N between the basic number of scans N0 and the actual number of scans N. If the moving speed is normalized by /N, the same character as that obtained when moving the sensor at the basic speed will be obtained (-0) Also, if the basic scanning number N0 is set to the maximum speed expected in manual feeding operation, If you set it to a value of 9, all the characters you get will be of a stretched size if you use the normal movement speed.To correct this.

All you have to do is thin out unnecessary information lines.

FIG. 6 shows an example of correction when N/No=2.

In the figure (α) is the basic number of scans N. 1 for the speed when
The character "'A" is shown read at a moving speed of /2.

In this case, the line number is '! , . . , l, 6, by thinning out odd numbers or even numbers as unnecessary information line numbers, the size can be corrected to the size for the basic speed as shown in FIG.

Even in the case of other moving speeds, correction can be performed by thinning out every fixed number of lines according to the value.

Further, when performing 9 thinning, a logical OR of the "black" level is performed between the line to be thinned out by 1 and the adjacent line, and the resulting line is used, thereby preventing deterioration in character quality due to thinning by 1.

FIG. 7 shows a circuit block diagram of this embodiment. In the figure, 9-1.9-2 is a one-dimensional image sensor, 10-1.
10-2 indicates an amplification/binarization circuit, respectively. 11 is a speed detection circuit, 12 is a memory circuit for one character, and 13 is a speed conversion ROM circuit.

14 is an output terminal. The photoelectric conversion output from the sensor 9-1 is converted into an image signal by an amplification/binarization circuit 10-1.

This shaped image signal is guided to a speed detection circuit 31 and a memory circuit 12 for one character. Similarly, sensor 9-2
The output is shaped into an image signal by an amplification/binarization circuit 10-2 and is led to a speed detection circuit 1. The speed detection circuit 11 calculates the moving speed ratio N/No based on both signals input from the sensors 9-1 and 9-2.

The speed conversion ROM circuit 13 receives the data calculated by the speed detection circuit 11, and sends the necessary thinning information for one character to the memory circuit 12. The memory circuit 12 for one character, which sequentially stores the photoelectric conversion output of the sensor 9-1, outputs to the terminal 14 an output corrected to the correct size as nine characters based on the thinning information from the speed conversion ROM circuit 13.

In this embodiment, an example in which the arrow 4 is moved to the right has been described, but the one-dimensional image sensor 9-1 is moved in the opposite direction to the left.
．． It goes without saying that if the detection in 9-2 is reversed, it is possible to read the same (-).Also, the size of characters, symbols, etc. written on the manuscript has been explained only within a relatively narrow range; By changing the size of the image sensor, it is possible to input images corresponding to the size of the image sensor, and it can also be applied to a manual facsimile.

(5) Effect description As explained above (-91-dimensional image sensors are installed at regular intervals, the movement speed of the sensor is detected, the ratio with the basic movement speed is determined, and the number of scanning lines is thinned out. By this, for any moving speed during manual feeding operation of the sensor,
It is possible to normalize read characters to a fixed character width. In addition, since an inexpensive one-dimensional sensor is used, the entire structure can be manufactured at low cost, making it suitable for simple OCR systems such as POS.
Alternatively, it can be applied to simple facsimile, etc.

[Brief explanation of the drawing]

FIG. 1 is a diagram of a reading device that performs manual feeding operation. Fig. 2 is a principle diagram of the reading method, Fig. 3 is an explanatory diagram of the reading state, Fig. 4 is an embodiment of the present invention and is a schematic diagram of the reading method, and Fig. 5 is a diagram of the reading operation in the embodiment of the present invention. An explanatory diagram, FIG. 6 is a diagram showing an example of the reading correction method,
FIG. 7 is a circuit block diagram of this embodiment. In the figure, 1 is the reading device, 2 is the original, 3 is the characters/symbols on the original, 4 is the movement direction, 5 is the reading signal line, 6 is the 2
dimensional image sensor; 7, lens optical system; 8, light source lamp; 9-1. ．． 9-2 is a one-dimensional image sensor, 10
-1.10-2 is an amplification/binarization circuit, 11 is a speed detection circuit, 12 is a memory circuit for one character, 13 is a speed conversion RO
M circuit. 14 indicates an output terminal. Patent Applicant Nippon Telegraph and Telephone Public Corporation Patent Attorney Mori 1) Hirosen 1 Me 6 O 4 Me (a) (b) Sai 5 Figure (a) (b) (c) ;A-3 [21 (a) ( b) 10 6 3 10 70

Claims (1)

[Claims] (1) In a system for reading characters, etc. written on a manuscript by hand-feeding operation, the first and second reading 1
Dimensional image sensors are arranged at regular intervals, and according to nine hand-feeding operations, the hand-feeding speed is detected from the timing difference between the first and second image sensors reading the same information, and the read signal of the image sensors is controlled. In the method. The maximum manual feed speed is set in advance, and the number of reading scans of the first and second image sensors required for this is set, and when the manual feed speed is slower than the maximum manual feed speed, the read information line obtained by scanning is An image reading method that is characterized by thinning out lines of unnecessary information. (2. In claim 1, the maximum manual feed speed is set in advance, and the first and second
Set the image sensor reading scanning number of 9 when the manual feeding speed is slower than the maximum manual feeding speed, among the reading information lines obtained by scanning. An image reading method characterized by performing logical OR thinning of an unnecessary information line and three adjacent read information lines.