JPH0245228B2 - - Google Patents

Description

TECHNICAL FIELD The present invention relates to a format control method for controlling an optical character reading device to read only a desired character area of a document.

(Background Art) Conventionally, when dealing with manuscripts in which formats such as character size, line spacing, and number of lines are specified, character lines can be determined by providing these as format information to an optical character reader. Reading control is being performed.

Furthermore, when the number of character lines in a desired character area is specified, reading of the desired character area can be controlled by adding line marks to each character line and counting them. .

(Problem to be solved by the invention) However, for example, when you want to read two character areas for the name and abstract of an academic paper, the font size and line spacing are often predetermined for each character area. However, the number of lines in each character area is generally undefined; for example, a name may have one line or two lines.

In order to specify each character area with an undefined number of lines, the present invention attaches a first mark to at least one end of the first character line of each character area, and also attaches a first mark to one end of the first character line of each character area. This is intended for documents with a second mark attached to the edge.

Further, the present invention provides a reading means for the first mark and an image sensor for reading characters and the second mark, and arranges the former in front of the latter with respect to the scanning direction. After detecting the mark, the last line of each character area is determined by starting writing the output of the image sensor into the memory and detecting the presence of a second mark based on the output of the image sensor, and then determining the last line of each character area. The first character line of the character area is detected, thereby controlling to read only the desired character area.

(Structure and operation of the invention) FIG. 1 is a block diagram showing an embodiment of the present invention. Here, the line pitch for each character area is known in advance, and the leading edge of the first character line of each character area is The target is a document with one line mark attached to the left edge of the document corresponding to the position, and one line mark attached to the right edge of the document corresponding to the last character line of each character area. This allows format control of the original size.

In FIG. 1, an image sensor 100 is provided with the left end of the document table as a field of view in order to read the left end line mark, and an image sensor 101 is provided with a field of view of the left end of the document table in order to read the left end line mark. The image sensor 100 is placed in front of the image sensor 101 in the scanning direction of the document so that the left edge line mark can be detected before reading the character line. The field of view in the document feed direction between 100 and 101 is made to have a difference equivalent to the maximum scheduled line pitch,
Image sensors 100 and 101 both read pixel lines.

The photoelectric conversion output of the image sensor 100 is given to the left end mark detection section 103, and when the left end mark detection section 103 detects the left end mark, the left end mark detection signal is sent to the format control section 104, and the format control section 104 detects the scheduled end mark. It detects the document feed corresponding to the maximum line spacing and gives an instruction to the write counter control section 105 to start writing, and further detects the document feed corresponding to the predetermined line pitch and instructs the write counter control section 105 to start writing.
5 to end the writing, and temporarily halts document feeding.

The write count control unit 105 controls writing in the row direction of the line memory 106 in synchronization with reading of each photoelectric cell of the image sensor 101,
Furthermore, writing in the column direction is controlled in synchronization with the vertical feeding of the document, so that an image for one character line is stored in the line memory 106 each time the feeding of one character line is completed. , and if a right end line mark is attached, that is also stored at the same time.

The read counter control unit 107 scans the address of the line memory 106 in the column direction, and reads the character pattern signal from the line memory 106 by scanning in the row direction every time one column scan is completed, and sends it to the character recognition unit. give. When the character recognition for one character line is completed, the format control unit 104 receives the completion signal from the character recognition unit, and sends the document again by the scheduled line pitch, and also stores the document in the line memory 1.
The stored information of 06 is updated.

The right end mark detection unit 108 detects the line memory 10.
The right edge line mark is detected from the read output of No. 6, and the document edge detection unit 109 detects the left edge of the document by checking whether the black background continues for a certain area or more. When the right end mark detection section 108 detects the right end line mark, the detection signal is sent to the format control section 104, and the format control section 104 determines that it is the last character line in the character area, and subsequently determines that the edge of the document has been detected. After confirming that the character area has ended, the process waits for the end of character recognition and then moves on to detecting the left end line mark of the next character area.

In this embodiment, the position of the right-most line mark is variable (the left-most line mark is fixed because it is aligned with the left end of the document table), and as a result, control that does not depend on the size of the document is possible.

Note that even when the target is a document with an undefined line spacing, the left edge line mark is attached not only to the first character line but to all character lines, and the image sensor 1 is sequentially attached each time the left edge line mark is detected.
By controlling the writing of the output of 01 and determining the detection of the first left end line mark from the edge of the document and the detection of the first left end line mark after the right end line mark is detected as the first character line of each character area. , it is possible to read only a desired character area, as in the previous embodiment.

(Effects of the Invention) In the present invention, since the mark indicating the last character line of each character area is read by an image sensor that reads characters, there is no need to attach the mark fixedly to the edge of the document or There is an advantage that the sizes of the two can be different.

Furthermore, since the first character line is detected before reading the characters and then stored in the memory, there is an advantage that reading unnecessary portions of the image can be avoided.

Furthermore, since only one type of left end mark is required,
The image sensor has the advantage of being a simple means with low resolution, such as a photocoupler.

[Brief explanation of drawings]

FIG. 1 is a block diagram showing one embodiment of the present invention. 100, 101...image sensor, 103...
...Left end mark detection section, 104...Format control section, 105...Write count control section,
106...Line memory, 107...Reading count control unit, 108...Right end mark detection unit.

Claims (1)

[Scope of Claims] 1. A manuscript in which a first mark is written on at least one end of the first character line of each character area with an undefined number of lines, and a second mark is written on the other end of the last character line of each character area. An image sensor 101 that reads characters and the second mark with a field of view wider than a predetermined maximum character line width in a format control method that controls to read only desired character areas. a memory 106 for storing the photoelectric conversion output of the image sensor; and a first mark reading means 100 for reading the first mark with a predetermined field of view for the first mark.
The first mark reading means is arranged in front of the image sensor with respect to the scanning direction of the document, and the first character line of each character area is read based on the photoelectric conversion output of the first mark reading means. After detecting the first mark, start writing the output of the image sensor to the memory, and start writing the output of the image sensor to the memory, and write the output of the image sensor to the memory.
The end of each character area is determined by detecting the mark, and the first character line of the first character line of the next character area is determined.
A format control method characterized by moving to mark detection.