JPS5835663A - Picture processing device - Google Patents

Abstract

PURPOSE:To edit a document picture with a simple instruction without providing a picture display part, by providing a table based on structure information in a memory. CONSTITUTION:An input original is made binary by a photoelectric conversion picture reader 201 and is transferred to a picture storage memory 202. Simultaneously, character strings are extracted from the picture by a structure information extracting part 203, and a table where respective character string line numbers and absolute positions in lines correspond to absolute coordinates of the memory 202 is generated and is stored in a table memory 205. This generated table is referred to designate line numbers or/and clause numbers, thus editing character strings. The editor watches the original to input the deleting or moving operation from a keyboard 208. When a line number and a clause number are inputted, a processing part 207 refers to the memory 205 to insert coordinates of the memory 202 corresponding to them and performs the editing for deletion, movement of a character string, expansion, or reduction and writes the result in a buffer memory 206.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention provides an O
This relates to A (office automation) equipment.

A conventional image processing device is shown in FIG. 1st and 1st ((1) is an input section, and 102 is an image storage memory.

103 is an image display section, 104 is a processing section, 105 is a buffer memory, 106a is an output section, and 101j is a keyboard. First, the original to be edited is transferred to the image storage memo +7 (102) by a photoelectric conversion image reading device. The image information is transferred to the image display section (103) and contacted by the editor. The editor looks at the screen, moves the cursor on one screen, and inputs 9 position information.

Input one more editing function. This editing operation is performed using the keyboard. The processing device (104) judges the position information and the input information of editing contents, and edits the image 1M on the memory (
(deletion, addition, enlargement/reduction), and output to the pack memory. The editor checks the editing results on the screen (103) and
Output when editing is complete.

Therefore, since the main editing operations are performed using the image display section and keyboard operations, there is a drawback that one image display section is required.

The present invention is intended to improve the above-mentioned drawbacks of the conventional technology, and aims to provide an image processing device capable of editing a document image with a simple command without the need for the conventional Q image display unit (103). consists of an input unit (201) that inputs data into the image storage memory (202) through photoelectric conversion of the original to be edited, a processing unit (207) that processes the input information, and a keyboard (207) for @ collection operations. 208) and an output unit (209), a structural information extraction unit (203) extracts a character string from one image information and associates the line number of the character string with the absolute coordinates of the representative point of the character string by the line number. ) and a table memory (205) for storing the output thereof, and all images corresponding to the line number and segment number specified from the keyboard (208) are searched and processed by referring to the table memory (205). It is found in image processing devices such as

This will be explained below with reference to the drawings.

Is Fig. 2 the first implementation of the present invention? 1] and 201 is an input section.

202 is an image storage memo IJ, and 203 is a structural information extraction unit.

204 is a buffer memory 2,205f/! A table storage memory, 206 a buffer memory 1, and 207 a processing unit.

208 is a keyboard, and 209 is an output unit. The structure information extracting unit o203 further includes an initial value setting unit 210, an arithmetic unit 211, a determining unit 212, and a table creation unit 213, as shown in FIG.

The following explanation will be based on documents. The input document is binarized (par 0) by a photoelectric conversion image reading device (201).
"White, Pa 1" Black] and transferred to the image storage memory (202). At the same time, a structural information extraction unit (203) extracts character strings from the image, and creates a table in which each character string associates the line number and relative position within the line with the absolute coordinates of the image storage memory. All representative examples of tables are shown in Table 1.

Margin below C) Each line is represented by the coordinates of two points, and the divisions within one line are as follows.

Identifies by number from the beginning of the line. Structural information extraction unit (20
3) first refers to the image storage memory (202)' and converts the image by representing it with the logical sum of (4X4) pixels and using the information of "O" and "1".Here, (512X 512) pixel t (128X 128) Converted to elements.Next, for each element, give the character string area, line spacing, or blank space probabilistically.In other words, calculate the number of 1''s in the 4 neighborhoods of each element, and if it is 0, it is 0.25 If it is 1 digit, it is 0.50. If it is 2 or more digits, it is 1.00 and given probabilistically. This is the initial value setting part (
210) and stores it in the buffer memory 2. Next, the calculation unit replaces the probability p of each element with the average value a of the four neighboring probabilities as pp/((1-pXl-a)+pp). This is processed in parallel using buffer memory 2. (2i3
)'s determination unit counts the number of elements whose probability is 0.9 or more or 0.1 or less, and determines whether 95% of all elements have been reached. If the number is less than 95, the string is returned to the arithmetic unit and processed repeatedly to extract the unprocessed character string until it reaches 95% or more. Next, the extracted strings are classified by relative position. The dimensions of each character string are determined, and if it is kI long, it is assumed that there is a strong correlation in the vertical direction, and if it is horizontally long, it is assumed that there is a strong correlation in the horizontal direction, and each character string is integrated into a line. Each integrated line is assigned a line number, and a table (Edict 1 table) is created in which the absolute coordinates of the representative points of each character string are associated with the line numbers. By referring to the table created in this way, document processing is performed by specifying the line number or line number and the number of divisions within the line number. The editor looks at the manuscript and deletes or moves the line by inputting it from all keyboards. When the line number and segment number are input, the processing unit (207) searches and deletes all the coordinates 1205 on the corresponding memo IJ (202), 2 moves or edits enlargement/reduction of character strings,
(206) Write the result to the buffer memory. Furthermore, input the entire second manuscript and enter the necessary trading place? It can be extracted and inserted into the previously processed memory.

As explained above, the first embodiment has the advantage that it does not require four image display units because it is not necessary to specify the position information of the two edited portions using accurate coordinates. Such means are effective because many document image processes can be solved by simple cutting operations. Also.

Since it is sufficient for the command given by the editor to give information on the document structure, an easy-to-use image processing device can be obtained.

Even when photos and graphs are mixed, by dividing the area and having a table with the corresponding absolute coordinates of the photo area, it is possible to easily perform cutting operations such as extraction and deletion of each area.

Since the present invention has a table based on structural information in the memory, it can also be used for editing by voice input when editing free format documents, handwritten documents, etc.

[Brief explanation of drawings]

FIG. 1 is a block diagram of a conventional image processing device, and FIG. 2 is a block diagram of an image processing device according to the present invention. FIG. 3 is a block diagram of the structural information extraction unit 203 in FIG. 2. 201: Input unit, 202: Image storage memory, 203: Structure information extraction unit, 204: Buffer memory 2. 205;
Table storage memory 206: Buffer memory l. 207: Processing section, 208: Keyboard, 2 (+9: Output section. 21O: Initial value setting section, 211: Arithmetic section 212: Judgment section, 213: Table creation section. Patent applicant Oki Electric Industry Co., Ltd.?F Patent application Representative Patent Attorney So Yamamoto - Go to Figure 1, Figure 2, Figure 3, 204

Claims (1)

[Claims] An input unit (201) that inputs the original to be edited into the fault storage memory (202) via full photoelectric conversion, a processing unit (207) that processes the input information, and a keyboard (207) for operating the M collection. 208) and an output unit (209), a structural information extraction unit (203) extracts a character string from one image information and associates the line number of the character string with the absolute coordinates of the representative point of the character string by vertical number. )4! :, has a table memory (205) that stores the output of
! : Image processing device 0 characterized by reference, search and processing