JPH0481071A - Digital copying device - Google Patents

Abstract

PURPOSE:To obtain an editing function, whose operation is comparatively easy, to make a fair copy at high level by writing recognized graphic information, information of connection between graphics and separated character information into a memory means as output pictures. CONSTITUTION:The picture information of a bit map inputted by a picture reading means (image scanner) 100 is written in a memory (means) 330 and afterwards divided into characters and graphics. The information divided as graphics is encoded for each small pattern such as a line segment, etc., and converted to vector information. Continuously, the relation of connection between recognized graphics is recognized, and the recognized graphic information, the information of connection between graphics and the information of first separated characters is prepared as output picture information. Thus, an output picture replaced with a registered clear graphic pattern and made into the fair copy is generated.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a digital copying device that can be used, for example, for making fair copies of handwritten drawings.

[Conventional technology] When creating drawings, in order to draw beautiful shapes and characters,
It is very time consuming as you have to use rulers etc. Therefore, if you want to shorten the time required to create one drawing, you will need to create the drawing by hand.

Professional tracers are often used to cleanly redraw handwritten drawings of this type, but this work is manual, time consuming, and expensive.

In departments with relatively large computer systems,
In some cases, the image information of a handwritten drawing is read with a digitizer or an image scanner, the image is input into a CAD system, and the drawing is edited and printed. However, even in this case, it takes a considerable amount of time to input and edit images, requires specialized knowledge in operating computers, and requires an expensive system.

[Problems to be Solved by the Invention] When using a digital copying machine, it is possible to output a hard copy of a drawing after performing editing such as moving the entire image on the document, cutting out areas, and overlapping the images. but,
Detailed editing processes such as fine-printing drawings cannot be performed. Also 1
For example, you may want to erase only text from an image in a drawing, but this type of operation is not possible with conventional copying machines.

Therefore, it is an object of the present invention to provide a digital copying device that is relatively easy to operate and has advanced editing functions such as fair copying at a low cost.

[Means for Solving the Problems] In order to solve the above problems, the present invention provides an image reading means for reading a document image as a set of minute pixel information, a memory means for storing the information read by the means, and an output means for outputting the image information stored in the memory means;
Figure separation means: Encoding means that encodes information on separated figures for each pattern such as a line segment: A figure that recognizes figures based on the connection relationships between encoded patterns, the number of patterns, and angles. Recognition means: A connection recognition means for recognizing connection relationships between a plurality of recognized figures; and the recognized figure information; connection information between nine figures; and separated character information, written in the memory means as an output image; output. image creation means;
will be established.

[Operation] According to the present invention, image information of a human map input by an image reading means (for example, an image scanner) is written into a memory means and then classified into characters and figures.

Information classified as figures is encoded for each small pattern such as a line segment, and converted into so-called vector information (for example, a line segment pattern is indicated by the coordinates of a starting point and the coordinates of an ending point).

Furthermore, the shapes represented by the patterns of the encoded information are recognized based on the connection relationships, number of clicks, and angles between the patterns of the encoded information. Next, the connection relationships between the recognized shapes are recognized,
Recognized graphic information 9 Connection information between the graphics and information on the first separated characters are created as output image information.

For example, when a handwritten drawing is input as a manuscript image, line segments and the like constituting figures in the input image are often curved, tilted, protruded, and the like. However, by registering the shapes made up of those line segments in advance and recognizing them as being the same as a specific shape pattern, it is possible to replace the deformed input shape button with the registered beautiful shape pattern. It is possible to generate a clean-cut output image using the following method.

By separating character information and graphic information in advance, it is possible to switch whether or not characters are output to an output image.

However, since it is very difficult to recognize a single figure, erroneous recognition of figures often occurs. If the erroneously recognized image is output as is, the contents of the input manuscript image and the fair drawing will be different, so the fair drawing cannot be used. Therefore, in a preferred embodiment of the present invention, which will be described later, it is possible to modify a recognized figure for a figure specified by input from a keyboard, area specification using a mark ben, or position specification using a coordinate input device. This correction function makes it possible to output correct clean-cut drawings even in the case of handwritten drawings that are difficult to recognize.

Other objects and features of the present invention will become clear from the following description of embodiments with reference to the drawings.

[Embodiment] FIG. 1 shows the structure of a mechanical section of a type of tendall copying machine that implements the present invention. Referring to FIG. 1, the second digital copying machine is roughly divided into an upper image scanner 10.
0 and a laser printer 200 placed below it.

A contact glass on which a document is placed is placed at the top of the image scanner 100, and an optical scanning system is provided below it. A document is exposed to light by an exposure lamp 1 of an optical scanning system, and its reflected light, that is, image light, passes through various mirrors and lenses 2 provided in the optical scanning system and forms an image on a light receiving section 3. This light receiving section 3 is provided with a one-dimensional CCD image sensor, which will be described later. Since the optical scanning system is driven by a mechanical drive system in the horizontal direction of the drawing, the image light obtained by exposing each part of the document surface is sequentially read by the image sensor, line by line.

The image information read by the image sensor.

The laser light is converted into an output image through processing described later and is output from the writing device 4 of the laser printer 200, which modulates the laser light. The laser beam modulated by the image information passes through a writing optical system and is imaged onto the surface of the photoreceptor tram 5. The entire surface of the photoreceptor tram 5 is uniformly charged to a predetermined high potential by the main charger 6 in advance, and when it is irradiated with image light, the potential changes depending on the light intensity, and the potential corresponding to the image is changed. distribution, that is, an electrostatic latent image is formed.

The electrostatic latent image formed on the photoreceptor drum 5 is made visible by adsorption of toner when it passes through the developing unit 7, forming a toner image.

On the other hand, recording paper is fed out from the paper feed cassette 12 or 13 selected, and the recording paper is fed to the photoreceptor drum 5.
The toner image is fed so as to overlap the surface of the photosensitive drum 5 in synchronization with the formation timing of the upper toner image. Subsequently, the toner image on the photosensitive drum 5 is transferred onto the recording paper by the urging of the transfer charger. Furthermore, by the energization of the charger 9 for one minute, the recording paper onto which the toner image has been transferred is transferred to the photoreceptor tram 5.
The image is separated from the camera and goes to the fixing unit 14. The toner image on the recording paper is fixed onto the recording paper by the fixing unit 14, and then the recording paper is ejected from the copying machine.

After the image transfer and recording paper separation are completed, the surface of the photoreceptor tram 5 is cleaned by the cleaning unit 10 in preparation for the next image formation.

FIG. 2 shows the configuration of the electrical equipment of the digital copying machine shown in FIG. 1. This will be explained with reference to FIG. image scanner 100
, the signal of the original image in the hint map format read by the CCD image sensor 0 is converted into a digital signal (8 signals in this example) by the A/D converter 120 and then sent to the shading correction unit 130.
Corrections were made for density level variations.

It is stored in memory unit 330. Memory unit 3
30 includes a storage area for one frame of one image, as well as an editing storage area and an output image storage area. The editing unit 340 includes the memory unit 3 as described below.
The input image information on 30 is processed to generate output image information. In this example, a figure dictionary 350 is provided in which characteristic parameters of various figures are registered, and the editing unit 340 uses this dictionary to recognize figures in image information.

The output image information generated on the memory unit 330 is
The signal is applied to the laser printer 200 in the form of binary information corresponding to black/white of each pixel, passes through the bumper 220, passes through the LD driver 230, and is applied as an energizing signal to the laser diode 240.

Therefore, the laser diode 240 outputs laser light modulated according to the output image information. This laser light is output from the writing device 4 and is irradiated onto the surface of the photosensitive drum 5 via a writing optical scanning system.

Instructions from the operator are carried out by keystrokes from an operation board 310 arranged on the top of the copying machine. The main control unit 320 controls various displays on the operation port 310, and also reads key input from the operation port 310 and instructs each section to start reading, 1 content of editing process, 1 image start copying, etc. .

FIG. 3 shows the contents of the operation of the copying machine shown in FIG. 1, mainly editing processing. First, an outline of each process shown in FIG. 3 will be explained.

In step 51, the original image is transferred to the image scanner 100.
The read image data is converted into binary information corresponding to black/white for each pixel and stored in the memory unit 330. In the next step 52, the memory unit 33
The image data stored in 0 is processed to separate character components and graphic components contained in the image data. For example, when an image as shown in FIG. 4a is read as a manuscript image, the process of step 52 separates two sets of graphic images shown in FIG. 4b and character images shown in FIG. 4c. Image data is obtained.

In the next step 53, the separated graphic component image data is processed and converted into vectorized data. That is, a continuous pixel group pattern is recognized as a line segment padder that approximates it, and a series of codes indicating each line segment, that is, each coordinate (x) of the starting point and ending point.

y) Obtain data. For example, when the graphic image shown in FIG. 4b is processed, a curved line segment is recognized as a straight line, so an image in which the curve of the line segment is corrected is generated as shown in FIG. 4d.

In the next step 54, the vectorized data representing a large number of line segments is processed, related line segments are grouped according to the connection status (distance between coordinates) between the end points of each line segment, and the line segments for each group are Recognize the shape of a figure from the connection relationship between the minutes (angle between line segments, number of line segments, etc.). When the vectorized data shown in Fig. 4d is processed, the image components shown in Fig. 4e, for example, are recognized as figures, and the remaining image components (line segment information) shown in Fig. 4f are connection lines and arrows between figures.

In the following step 55, each line segment is converted into a connecting line shown in FIG.
Each is recognized and classified as an arrow shown in figure h.

In the next step 56, an output image is selected according to a mode specified in advance by a key operation by the operator.

The output image mode includes a mode that includes characters and a mode that deletes characters. When a mote that includes characters is specified, the image information of the character components separated in step 52 is stored in the memory unit 330. Write to the output image area above.

When the mode does not include characters, image information for one character is not written to the output image area.

In the subsequent step 57, vector data such as recognized figures and connection lines are aligned horizontally and vertically (correcting the slope of line segments, protruding end points, etc.), and the opening angles and lengths of arrows are aligned. Thereafter, it is converted into bitmap information and written into the output image area on the memory unit 330. As a result, output image information as shown in FIG. 41 or FIG. 4j is generated in the output image area.

In step 58, the information on the output image created in the processes up to step 57 is output to the laser printer 20O to create a hard copy of the corrected image. This completes the fair copying of the handwritten manuscript drawing.

Although only the outline of the process is shown in FIG. 3, in reality there is a possibility that the figure may be misrecognized in step 54, so if there is a misrecognition, correction mote processing is executed. . If a graphic error is found on the hard copy of the output image output from the printer 200, the correction mode is entered by pressing the correction mode key on the operation board 310. For example, if the image shown in FIG. 5a is output in response to the input image shown in FIG. 4a, there is a misrecognition of the figure, so it is necessary to execute the correction mode. In the modification mode, first a message is displayed on the display of the operation board 310 asking you to specify the position of the figure as shown in Figure 5b, so enter the number to specify the position using the numeric keypad. do. In this embodiment, each shape is numbered in order from left to right and from top to bottom, so for example, the misrecognized shape in FIG. 3. Next, as shown in Figure 5c, a message will be displayed on the display of the operation boat 310 asking you to specify the number that identifies the correct shape, so enter the number that has been assigned in advance to the correct shape using the numeric keypad. do. In this example, a horizontally long rectangle is registered as shape number 1 in advance, so when you enter 1 from the numeric keypad, the erroneously recognized hexagonal shape is changed to a horizontally long rectangle as shown in Figure 5d. replaced by After this, when the modification mode is terminated, an output image is again created on the memory unit 330 based on the modified information, and this image is hard-copied by the laser printer 200.

Next, the specific contents of each step in FIG. 3 will be explained.

Step 52: Trace the outlines of sets of black pixels adjacent to each other, and determine the size (width and height) of the outline of each set. Then, the obtained size is compared with a predetermined threshold value (fixed value), and based on the result, it is determined whether the pixel set is a character or a figure.

Contour tracking is processed as follows. First, the input image is rask-scanned to find a black pixel to start tracking.

Then, from that black pixel, contour pixels are traced counterclockwise in the case of an outer contour line, clockwise in the case of an inner contour line, and when it returns to the position of the tracing start pixel, that pixel set is traced. finish. This process is repeated until there are no more pixels for which tracking has not been completed. An example of tracing the outline of one black pixel set is shown in FIG. 6a. In FIG. 6a, pixels subjected to hunting indicate black pixels.

Other pixels indicate white pixels. In this example, eight types of directions 0 to 7 were used as the directions of the contour line, as shown in FIG. 6b. Referring to FIG. 6a, a rask scan is first performed as shown in 30 to search for a tracking start pixel. In this example (i+
1. The tracking start pixel is found at position j), and since the previous pixel in the rask scan is a white pixel, it is regarded as the outer contour line and tracking is started counterclockwise from this position.

Next, neighboring pixels are checked counterclockwise from direction 4 in FIG. 6b, and the direction of the first black pixel found is set as the direction of the contour line. Next, the center pixel of tracking is moved to that black pixel, neighboring pixels are checked counterclockwise from the direction of the previous contour line -2, and this is repeated until the tracking start pixel is reached. As a result, a contour tracing result as shown by the arrow in FIG. 6a is obtained.

Among the tracing results, only the outer contour line is stored in the memory unit 330 as contour data. In this example, the contour data is expressed as starting position (minimum) and length (maximum, minimum) values based on the minimum and maximum coordinates of the contour, as shown in Figure 6c. Ru. In the example of the black pixel set in FIG. 6a, 1. As can be seen from the contour data in Figure 6c, the X coordinate of the starting position is 1. The X coordinate is J
.

The length in the X direction is 5 (pixels) and the length in the y direction is 4 (pixels).
It has become.

Of the contour data obtained as a result of the above tracking, x, 37
Based on each length in the direction, the size of the black pixel set is determined, and it is determined whether the image is noise, a single character, or a figure. In other words, the upper limit L1 of the size of the noise image and the upper limit L2 of the size of the character image are determined in advance, and as shown in FIG. 7a, the lengths X and Y of each axis data are both smaller than Ll. If the lengths Coded into emotional sounds that indicate the classification of letters and noise,
The information is recorded in the form shown in FIG. 7b in an order corresponding to the arrangement of the surplus data (FIG. 6c).

However, when reading the original with an image scanner.

When reducing or enlarging an image, the size of each image is different from the normal size, so the threshold value Ll' and L2' corrected as follows based on the reading magnification R are used for judgment. do.

L1'= LI X R L2' = L2 As disclosed in No. 4 (pages 845-852) and Japanese Patent Application Laid-open No. 1986-286177, the center line is determined while tracing the contour lines on both sides of the line figure image, and this is converted into a polyline hector. An approximate representation method is used.

Step 54 (1) First, it is checked whether or not another line segment is connected near the end point of each line segment, and unconnected line segments are removed.

This process is repeated until there are no unconnected line segments. For example, when processing the line segment data of the image shown in FIG. 7C, the line segments indicated by ■ are removed in the first processing,
In the second process, the line segment indicated by ■ is removed, and in the third process, the line segment indicated by ■ is removed, and the process ends.

(2) Among the line segments remaining after the above processing, the same line segment is
Find the smallest loop without going through it twice. When a branch occurs, priority is given to the direction in which the sign of the angle between the line segments is the same. Returning to the starting point completes the extraction of one loop. This process is repeated to extract line segment groups that are candidate figures. As a result, line segment groups as shown by ■ and ■ in FIG. 7d, for example, are detected.

(3) For the line segment groups extracted by the above process, compare their feature quantities (number of line segments, length ratio, angle, etc.) with the feature quantities of various figures registered in the figure dictionary 350, and A figure with similar quantities is recognized as the figure indicated by the line segment group. If a figure with similar quantity of features cannot be found, the process returns to the above process (2) and searches for another loop.

The remaining line segments that were not recognized in Step 55 and Step 54 are: ■ Connecting lines between figures (connected at both ends), ■ Line segments connected to figures (connected at one end), ■ Isolated line segments (not connected at both ends), and ■Arrow head (two short line segments are classified into four types: long line segments and connections. This classification is performed by the following first-order processing.

(1) Check whether there is a figure connected to the end of the line segment. With this, 1.(D, <2'l, ■) can be determined.

(2) Regarding the line segment marked ■ or ■ determined in (1) above,
Determine whether it is the head of the arrow.

Next, another example will be described. In this embodiment, most of the configuration and operation of the device are the same as those in the previous embodiment, but the correction process when a figure is misrecognized is changed. That is, as shown in FIG. 8b, if an erroneous recognition of a figure as shown in FIG. 5a is found in the clean-up output (hard copy) of the image, the operator may check the hard copy as shown in FIG. Then, the misrecognized shapes are surrounded by marks with predetermined characteristics.

Mark the area. The marked hard copy is read as a document by an image scanner 100. In this image reading, the position of the marked area is identified. Marking Ben's ink has a lower density than regular images.

Since its density is approximately constant, the image data before being binarized is processed and the marked area is detected by identifying at. A figure located within this area is extracted as an erroneously recognized figure from the figure data group that was the source of the previous output image. When a number specifying a correct graphic pattern is input from the operation board, the erroneously recognized graphic is replaced by the V shape of that number. Thereafter, an output image is created again for the image with the recognized mouth shape corrected, and the output image is copied 71 by the printer 200.

When there are multiple misrecognized figures, it is necessary to mark multiple areas and specify multiple positions. In this case, when inputting the correct figure number, the position information of each area (for example, the center position of the area) is displayed on the display on the operation board as shown in Figure 8C, and the current modified figure number is displayed. Make sure you know the location. Note that instead of displaying the numerical value indicating the position, the figure to be corrected may be displayed as shown in FIG. 8d.

Further, both the designation of the position of the erroneously recognized figure and the designation of the number specifying the correct figure may be performed by marking. That is, as shown in episode 9a.

Write a box specifying the area using Mark Ben, and then write a number in Mark Pen in the box.

In this case, if the marked copy is read with an image scanner, the area designation information and figure number designation information are automatically identified from the marked image, and correction processing is executed based on them, the operator can make the corrections. No key manual effort is required (see Figure 9b immediately).Identification of the mark image in this case can be carried out, for example, by the process shown in Figure 9C. In the process of FIG. 9C,
First, after tracing the contours of the read mark image, the inclusion relationship (inside or outside) is determined based on the position and size of each contour.If it is outside, it is determined that the area is specified, and if the The figure data is regarded as an incorrectly recognized figure. If the determination result is inside, it is determined that the figure number has been specified, and that part of the image is cut out and character recognition is performed to detect the correct figure number. If the figure number cannot be recognized, you can switch to number designation using the numeric keys on the operation board.

Mark images can also be identified by determining the size of one contour, as shown in the process of FIG. 9d. That is, if the size of the result of contour tracing is equal to or larger than a predetermined threshold value, it is regarded as area designation, otherwise it is regarded as graphic number designation. Other processing is the same as in the case of FIG. 9c.

In addition, as shown in Figure 1O, if two types of marking pens with different ink densities are prepared and one is used to specify the area and the other is used to specify the figure number, it is possible to identify the image density. This makes it possible to distinguish between area designation and graphic number designation without any special processing.

The mark image identification process in this case is shown in FIG.

Next, another embodiment will be described. In this embodiment, in order to correct erroneously recognized figures, as shown in FIG. A device is provided in which the devices are overlapped and integrated. A stylus bar for specifying a position is connected to the tablet device.

FIG. 13 shows the contents of the process related to correction of misrecognized figures. For example, if a misrecognition as shown in Figure 5a occurs, a portion of the output image is displayed on the display device, and the operator can scroll the displayed image by operating a predetermined key to see the entire image. While looking through the tablet device at the misrecognized figure, move the stylus pen on the tablet so as to surround it (see Figure 14a). As a result, the position of the figure to be corrected is read, and the figure at that position is specified as the object to be corrected. Thereafter, use the numeric keypad to input the number of the correct figure corresponding to the incorrectly recognized figure.

Once the modifications have been completed, an output image is again generated, which is then hard-copied by the printer.

It is also possible to specify both the position of the misrecognized figure and the number of the correct figure by operating the stylus pen. In other words, as shown in Figure 14b, move the stylus so as to surround the misrecognized shape, and then move the stylus so that it forms the shape of the number (1 in this example) that identifies the correct shape. do. In case 2, the processing is performed as shown in Figures 15a[i> and 15b, so there is no need to operate the numeric keypad to input the number.

Note that the correspondence between a graphic number and a graphic shape may be displayed on the display device.

One modified example will be described. In this example,
Using a device similar to that shown in Fig. 12, the position of the misrecognized figure is specified with a stylus pen as shown in Fig. 14a, but the operation for specifying the correct figure is performed on that figure as shown in Fig. 15c. This is done by directly inputting the shape of the object using a stylus pen. The shape entered with the stylus pen is compared with the feature parameters in the shape dictionary, and the returned shape is recognized. The misrecognized figure is then replaced by the recognized figure. The contents of the modification operation in case 2 are shown in Figures 5 and 5d, and the contents of the modification process are shown in Figure 15e.

Processing of information input manually using a stylus pen will be explained. As shown in Figure 17, coordinate data is sampled at predetermined intervals in response to input from the stylus pen, and the angular error (
The number of points is reduced by approximating the line so that the distance error (Fig. 16b) when connecting the current point and the point two points ahead (Fig. 16a) is less than a predetermined threshold. At the same time, distinguish between straight parts and curved parts.

[Effects] As described above, according to the present invention, the following effects can be obtained.

Claim 1: When inputting a handwritten drawing as a manuscript image, each line segment or the like constituting a figure in the input image has one bend.

In many cases, it may be tilted or protrude.

By registering the shapes made up of those line segments in advance and recognizing them as being the same as a certain specific shape pattern, the deformed input shape pattern can be replaced by the registered clean shape pattern, thereby erasing it. It is possible to generate an i-imaged output image.

Claim 2: Since the character information and the graphic information are separated in advance, it is possible to easily switch between outputting characters on the output image and whether or not characters are to be outputted in response to an instruction from the operation board.

Claim 3: Even if a figure is misrecognized, it can be corrected and a correct image can be output.

Claims 4 to 7: Since the position of the erroneously recognized figure is specified by marking, the correction operation is easy.

Claim 8: Since the output image is displayed by the display means before actually making a hard copy, if there is a misrecognition of a figure, the hard copy can be made after corrections are made, thereby eliminating unnecessary copies. The frequency of occurrence can be reduced.

[Brief explanation of the drawing]

FIG. 1 is a front view showing the structure of a mechanical section of a digital copying machine according to an embodiment. FIG. 2 is a block diagram schematically showing the electrical equipment section of the apparatus shown in FIG. 1 in horizontal lines. FIG. 3 is a flow chart showing an outline of the operation of the apparatus shown in FIG. Figure 4a, Figure 4b, Figure 4c, Figure 4d, Figure 4e,
Figures 4f, 4g, 4h, 41, and 4J
The figure is a plan view showing the contents of an input image or a whole or part of an image undergoing editing processing. Figures 5a, 5b, 5c, and 5d are images in which misrecognition has occurred, the appearance of the operation board when specifying the position to be corrected, and the appearance of the operation board when specifying the correct figure, respectively. , and a plan view showing the corrected image. Fig. 6a is a plan view of an image to be processed showing an example of contour tracing processing, Fig. 6b is a plan view showing divisions in the direction of contour tracing processing, and Fig. 6c is the content of contour data obtained as a result of the processing. This is a memory map showing the Fig. 7a is a flowchart showing the process of character 1 figure and noise identification, Fig. 7b is a memory map showing the arrangement of data obtained as a result of identification, and Fig. 7c is an example of a line segment pattern when extracting figure candidates. A plan view, FIG. 7d is a plan view showing an example of a line segment pattern at the time of loop extraction. Figures 8a and 9a are plan views showing a portion of the marked image, Figures 8b and 9b are flowcharts showing the operating procedure for outputting a fair copy in each embodiment, and Figures 8c and 8d are FIG. 3 is a plan view showing the appearance of the operation board in a correction mode. FIGS. 9c, 9d, and 11 are flowcharts showing processing for marking images, and FIG. 10 is a graph showing the density of general images and marks. FIG. 12 is a perspective view showing the appearance of the device of one embodiment;
Figures 15a and 15d are flowcharts showing the operating procedure for outputting a fair copy; Figures 14a, 14b and 15c are plan views showing the displayed image and the movement locus of the stylus pen thereon; Figures 15b and 15e
The figure is a flowchart showing the procedure of correction processing, Figures 16a and 16b are plan views showing the relationship between sample knob 3, ←, angle error and distance error, and Figure 17 is the movement trajectory of the stylus pen and the processing result. FIG. 1 Exposure lamp 2 Lens 3 Light receiving unit 4 Writing device 5, g light tram 6°, main charger 7 Developing unit 100° Image scanner 110: CCD image sensor 120: A/D converter 200 Laser printer 310
゛Operation board 330 Memory unit 340 Editing unit 3500 Graphic dictionary Treasure map funeral map Section 4a Voice to 44 Human power Section 4b Section Main scanning direction Fig. 7b Judgment results Fig. 7d Loop extraction Fig. 9a Correction (position and correct figure number) Fig. 9b Correction flow (2) Fig. 8a Correction (position specification) Fig. 8C Correction (figure number input) Fig. 8b Correction flow Figure 8d Correction (figure number manual) Figure 9C Judgment flow of area and number Voice] Figure tablet + Graphic display area Figure 13 Operation flow for correction Figure 14a Correction (position specification) Stylus pen trajectory Figure 14B Correction (correct figure number) Correct figure number Figure 15A Operation flow for correction (2) Section 15c Figure correction (correct figure 11B shape) Figure 15b processing flow Figure 15d operation flow for correction (3) Section (a) Stylus pen trajectory Figure 17 (b) Sampling point (C1 Linear approximation

Claims (8)

[Claims] (1) A digital copying device that includes an image reading means for reading a document image as a set of minute pixel information, a memory means for storing the information read by the means, and an output means for outputting the image information stored in the memory means. In: Character/figure separation means for separating input image information stored in the memory means into characters and figures; encoding means for encoding information on the separated figures for each pattern such as a line segment; A figure recognition means for recognizing a figure based on the connection relationship between each of the recognized patterns, the number of patterns, and an angle; A connection recognition means for recognizing the connection relationship between a plurality of recognized figures; and the recognized figure information, the figure A digital copying apparatus comprising: an output image creation means for writing connection information between the characters and separated character information into the memory means as an output image. 2. The digital copying apparatus according to claim 1, further comprising: (2) selection means for selecting whether or not to register the separated character information in the output image in accordance with an instruction. (3) The digital copying apparatus according to claim 1, further comprising a correction means for correcting the result of figure recognition for a figure at a designated position. (4) The correction means determines the position of the specified figure by identifying a predetermined image density.
The digital copying apparatus according to claim 3. (5) The correction means detects the contour pattern of a region with a predetermined image density, and automatically identifies whether each contour pattern indicates a region or not based on the inclusion relationship of the plurality of contour patterns. 4. A digital copying apparatus according to claim 3. (6) The correction means detects contour patterns of a region of a predetermined image density, and determines whether each contour pattern indicates a region or a character pattern depending on the size of the detected contour pattern. 4. The digital reproduction apparatus of claim 3, wherein the digital reproduction apparatus automatically identifies. (7) The correction means identifies at least two types of predetermined specific image densities, and the first specific image density and the second specific image density.
4. The digital copying apparatus according to claim 3, wherein whether the designated pattern in the image is an area designation or a character designation is automatically identified based on the distinction between the specified image density and the specific image density. (8) A display means for displaying two-dimensional image information and a two-dimensional position information input means, the modification means displaying the position specified by the two-dimensional position information input means together with the image on the display means, 4. The digital copying apparatus according to claim 3, wherein the result of figure recognition is corrected with respect to the figure at the position.