JPH09128429A - Method for recognizing dimensional numerical value and its device - Google Patents

Abstract

PROBLEM TO BE SOLVED: To accurately and automatically recognize dimensional numerical value from a drawing with dimensional information. SOLUTION: The picture of a drawing is read and its image picture information inputted. Each extention line within the inputted picture information is extracted to search dimension lines existing between the extracted extension lines to search which side area of the extracted dimensional line the character exists on. Then, character recognizing processing is executed to an area near the dimensional line on the side judged to have the character, and between the dimension lines to recognize the dimensional numerical value.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention recognizes dimensional numerical values in drawings having dimensional information consisting of auxiliary dimension lines, dimensional lines, and numerical values, such as construction drawings such as architectural drawings and various design drawings. The present invention relates to a dimensional numerical value recognition method and apparatus.

[0002]

2. Description of the Related Art In recent years, using CAD systems, etc., construction drawings including houses, buildings, etc., construction drawings including water pipes, gas pipes, piping drawings for power / communication cables, etc. can be easily created, and the data thereof can be obtained. It has been practiced to memorize and store the information when design changes, extension and remodeling are performed. However, the data of the construction drawings are not compatible with each other due to the created system, and cannot be used due to the passage of time or the change of the contractor. In addition, when renovating or remodeling a house, there are many cases where there is only an old architectural drawing drawn on paper, and it has been necessary to redraw all parts including the floor plan of the renovation and remodeling, etc.

Therefore, it has been attempted to read construction drawings such as architectural drawings drawn on paper and recognize and store them as computer processable data, but there is no special method or device for this purpose, A drawing is read by an image scanner, the image data is input to a personal computer or the like, and line segment recognition and pattern recognition are performed using a general figure recognition function. Alternatively, the graphic recognition function can be upgraded by using a high-performance graphic editor to assist, and even if the automatic recognition function is somewhat imperfect, basic vector diagrams such as straight lines and arcs can be converted by raster-to-vector conversion, for example. Some of them can be automatically recognized.

[0004]

However, in such construction drawings and various design drawings, dimensional auxiliary lines,
Many dimensional information consisting of dimensional lines and numerical values is described. This is important information indicating the actual size of each part drawn in the drawing, but the conventional drawing recognition device cannot automatically recognize this size information, especially the size value. Therefore, the operator manually inputs the numerical values one by one from a keyboard or the like.

Even when the dimensional numerical value is read by using the OCR, it is necessary to specify each reading area in which the dimensional numerical value is described. Therefore, when the contents of the drawings having the dimension information are converted into a database that can be processed by a computer, there is a problem that the input of the dimension values is complicated and the input error may occur in the conventional technique.

The present invention has been made in view of the above points, and it is possible to automatically and accurately and quickly recognize a dimensional value which is important information from image information of a drawing having dimensional information such as a construction drawing. The purpose is to do so.

[0007]

In order to achieve the above object, the present invention provides a dimensional numerical value recognizing method and a dimensional numerical value recognizing apparatus in the following drawings. A method of recognizing dimension numerical values of a drawing according to the present invention inputs image information of a drawing having dimension information including a dimension auxiliary line, a dimension line, and a dimension numerical value, extracts each dimension auxiliary line in the input image information, The dimension lines existing between the extracted dimension extension lines are searched, the area on the side where there is a character with respect to the found dimension line is searched, and the dimension is determined near the dimension line on the side where it is determined that there is a character. It is characterized in that character recognition processing is performed on a region between lines to recognize a numerical value.

Further, the drawing numerical value recognizing device according to the present invention includes image information input means for inputting image information of a drawing having size information consisting of auxiliary dimension lines, dimension lines and numerical values, and the inputted image information. Means for extracting each of the dimension auxiliary lines, means for searching the dimension lines existing between the dimension lines extracted by the means, and which side of the searched dimension line the character is in And means for recognizing the dimension numerical value by performing character recognition processing on the area between the dimension extension lines in the vicinity of the dimension line on the side where the character is determined to be present by the means. .

Therefore, it is possible to automatically search the area in which the dimensional numerical value is described and recognize the dimensional numerical value by OCR or the like, so that the dimensional numerical value can be manually input or the reading areas can be individually read. There is no need to specify it.

Further, in the dimensional numerical value recognizing apparatus of this drawing, the data of the dimensional line which can be searched, the data of the two dimensional auxiliary lines corresponding to the dimensional line, and the dimensional numerical value recognized about the dimensional line are put together. It is advisable to provide dimension list creating means for creating the number of dimension lists described above according to the number of dimension lines.
As a result, the numerical values and the dimension lines can be associated with each other on a one-to-one basis.

Further, it is preferable to provide a calculating means for calculating a dimension numerical value related to another dimension line in the vicinity from the already recognized dimension numerical value based on the information of the dimension list created by the dimension list creating means. As a result, it is possible to calculate the dimension values of other dimension lines in the vicinity from the recognized dimension values without performing all the dimension values by character recognition processing such as OCR, thereby reducing the recognition time and the recognition accuracy. Can be improved.

If the image information input means blade is an image reading means for reading an image of a drawing and inputting the image information of the image, it is possible to easily recognize the dimensional numerical value on the drawing drawn on the paper. . Further, if the image information input means is an image information receiving means for receiving and inputting the image information of the drawing by communication, the image information of the drawing is received from another device at a remote place, It is possible to easily recognize the dimensional numerical value of.

Further, by providing a display means for displaying the recognized dimensional numerical value at the same time as the input image information image, it is possible to easily confirm the recognition result and find the erroneous recognition. If the recognition result confirmation means for confirming the recognized dimensional numerical value displayed on the display means by an instruction from the outside is provided, the operator can confirm the recognition result and then confirm it by the instruction.

[0014]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be specifically described below with reference to the drawings. FIG. 1 is a block diagram showing a schematic configuration of a dimensional numerical value recognizing apparatus for carrying out the dimensional numerical value recognizing method according to the present invention, in which a hardware configuration and a software processing function by a microcomputer are shown in a mixed manner.

This device comprises an overall control unit 1, an image reading unit 2, a communication control unit 3, a memory 4, an automatic skew correction unit 5, and
It is composed of a dimensional position recognizing unit 6, a dimensional list creating unit 7, a dimensional numerical value recognizing unit 8, a display unit 9, an operation input unit 10, an external storage device 11, a printing device 12 and a bus 13 connecting them. In addition, these parts (or devices) and the bus 13
The interface part required between the and is not shown.

The overall control unit 1 is a microcomputer (CPU, RO) for controlling the operation and function of the entire apparatus.
M, RAM, etc., but is abbreviated as "CPU"), which is an automatic skew correction unit 5, a dimension position recognition unit 6, a dimension list creation unit 7, and a dimension value recognition unit according to the present invention. Each function of 8 can also be realized by the software processing of the CPU.

The image reading section 2 is an image information input means for scanning a drawing having set dimension information, reading the image, and inputting image information. The image reading section 2 includes a scanning optical system and an image sensor such as a CCD. It is a known image scanner including its drive circuit and the like. Also,
It also includes a binarizing circuit that binarizes the read image image information at a predetermined resolution to produce image data of white dots and black dots.

The communication control unit 3 receives the image information from the image reading unit 2 instead of receiving the image information from the image reading unit 2, and receives and inputs the image image information or the encoded image data in which the run length is encoded by communication from the outside. In addition to the means, the dimensional numerical value recognized by this device or dimensional information including it can be transmitted to an external device. Specifically, it includes a FAX modem and personal computer communication control means.

The memory 4 includes image image information read by the image reading unit 2, image image information received by the communication control unit 3 or encoded image data, image data skew-corrected by the automatic skew correction unit 5, and dimensions. A large-capacity RAM or a memory such as a hard disk for storing the dimensional area recognized by the position recognizing unit 6, the information of the dimensional list created by the dimensional list creating means 7, the dimensional numerical value recognized by the dimensional numerical value recognizing unit 8, and the like. is there.

The automatic skew correction unit 5 adjusts the angle of the image data stored in the memory 4 so that the horizontal and vertical line segment directions coincide with the horizontal and vertical reference directions (X and Y directions) of the apparatus. This is for correction, and a known automatic skew correction technique can be used. The image data corrected by the automatic skew correction unit 5 is stored in the memory 4 again.

The dimensional position recognizing section 6 has an area (hereinafter, referred to as "dimension area") in which the dimensional information in the image data exists for the image image data or the encoded image data which is automatically skew-corrected and stored in the memory 4. )), And a dimension auxiliary line is extracted for each of the dimension regions to recognize the dimension position.

The dimension list creating section 7 includes a dimension position recognizing section 6
The dimension lines between the dimension extension lines extracted by are searched, and the data of the extracted dimension lines, the data of the two dimension extension lines corresponding to the dimension lines, and the dimension values related to the dimension lines (dimensions described later) Dimension list creating means for creating a dimensional list in which the numerical value recognizing unit 8 is recognized.

The dimensional numerical value recognizing unit 8 searches for the area on the side of each dimensional line of the dimensional list created by the dimensional list creating section to find a character, and determines the size of the side on which the character is found. OCR for the area between the dimension lines near the line
Is a dimensional numerical value recognizing means for recognizing a dimensional numerical value by performing character recognition processing according to. Further, it also has a function of a calculating means for calculating the dimension numerical values of other dimension lines in the vicinity from the already recognized dimension numerical values based on the information of the dimension list. Details of these processes of the dimension position recognition unit 6, the dimension list creation unit 7, and the dimension value recognition unit 8 will be described later.

The display unit 9 receives the image data of the drawing which is input from the image reading unit 2 or the communication control unit 3 and is skew-corrected by the automatic skew correction unit 5, the dimension area and the dimension auxiliary recognized by the dimension position recognition unit 6. It is for displaying a line, a dimension line searched by the dimension list creating unit 7, a dimension value recognized by the dimension value recognizing unit 8, and the like, for example, a CRT or a liquid crystal display. In particular, it has a function of displaying the dimensional numerical value recognized by the dimensional numerical value recognizing unit 8 or the dimensional information including the dimensional line and the dimensional auxiliary line simultaneously with the image of the image information of the inputted drawing.

In this case, the recognized dimension information and the dimension information of the image of the inputted image information of the drawing can be displayed adjacent to each other or can be superimposed and displayed.
Then, in order to easily identify the two, it is preferable to display the image image in halftone and display the dimension value of the recognition result in high contrast. In the case where the display unit 9 is a color display device, the distinguishability can be improved by displaying the two images in different colors. Further, when the possibility of erroneous recognition of the recognized dimensional numerical value is high (equal to or more than a predetermined value), it is possible to display that the erroneous recognition is high by blinking the dimensional numerical value.

Further, the display unit 9 also displays a screen for the operator to confirm the recognition result. That is, the dimension value of the recognition result is displayed, and a display such as "Are you sure with this recognition result? (YES / NO)" is displayed. As a result, the operator can confirm whether or not the dimensional numerical value can be accurately recognized.

The operation input unit 10 is for inputting various operation instructions, function selection instructions, edit data, etc., and is a keyboard, a mouse, a touch panel, or the like. The operation input unit 10 also has a function as a display selection unit,
The display state of the display unit 9 can be changed to a display state desired by the operator. For example, by operating a key, it is possible to superimpose and display the input image image of the drawing and the dimension information of the recognition result, or select and display only one of them.

Further, the operation input unit 10 inputs a key or the like for inputting "YES" or "NO" information in response to the above-mentioned "is this recognition result OK? (YES / NO)" display. We also have means. Then, when "YES" is selected, the displayed recognition result is confirmed and the recognition process is ended, and when "NO" is selected, the process is shifted to the re-recognition process or the correction process. As a result, the operator can confirm the content of the recognition result and confirm it.

Further, the operation input unit 10 is provided with a mouse as a pointing device, and the dimension auxiliary line or the dimension line of the input image image of the drawing displayed on the display unit 9 is traced by the mouse cursor. It is also possible to provide a function for inputting position information of the or dimension line.

The external storage device 11 stores the dimensional numerical values recognized by the dimensional numerical value recognizing section 8 and the information of the dimensional list created by the dimensional list creating section 7 into a floppy disk (FD), a magneto-optical disk (OMD) or the like. Is a storage device for storing in a storage medium that can be taken out of the computer. The printing device 12 is a printer or plotter that prints or draws the input image image of the drawing or the recognized dimension information on paper and outputs it.

Here, the dimension information of the drawings will be described.
Generally, the dimension information is a description indicating the length, size, position, angle, etc. on the drawing. Construction drawings (mainly architectural drawings)
In Fig. 2, the dimension information indicating the length is usually represented by a combination of dimension figures as shown in Fig. 2. That is, a dimension line is drawn between two dimension extension lines that are drawn out from both ends of a portion where the dimension of the target graphic is desired to be designated, and a dimension value is written in the central portion on the dimension line. . That is, the dimensional values and the dimensional lines are in one-to-one correspondence.

FIG. 2A shows an example in which the dimension information is arranged on the upper side of the figure, and FIG. 2B shows an example in which the dimension information is arranged on the lower side of the figure. Although the position of the dimension line with respect to the dimension extension line is different for both, there is a dimension numerical value on the dimension line (on the right side if the dimension line is vertical), and both are closer to the dimension line below the number. . However, in some cases, the dimension number is below the dimension line (left side if the dimension line is vertical).
It may be described in.

FIG. 3 is a display example in the case where a long dimension is divided into a plurality of parts and the individual dimensions are shown.
When the dimensional information shown in is decomposed, it becomes as shown in (b).
FIG. 4 shows an example of a floor plan of an architectural drawing having actual dimension information.

Next, the procedure of the dimensional numerical value recognition process of the construction drawings (mainly architectural drawings of houses, buildings, etc.) by the dimensional numerical value recognizing apparatus shown in FIG. 1 will be described with reference to the flowcharts of FIGS. 5 and 6. To do. In the flowchart of FIG. 5, each step is indicated by “S”. In the example described here, the construction drawing to be recognized is assumed to be a positive drawing.

FIG. 5 shows the main routine of the processing relating to the present invention by the overall control unit (CPU) 1 of the dimension numerical value recognition apparatus for construction drawings according to the present invention. First, step 1
At, the image of the construction drawing set in the image reading unit 2 is read, and the image data obtained by binarizing the image is input and stored in the image data storage area of the memory 4. Alternatively, the communication control unit 3 may externally receive and input image data of the construction drawing (image image data or encoded image data obtained by encoding the run length).

In step 2, the automatic skew correction unit 5 makes the horizontal and vertical line segment directions of the input image data stored in the memory 4 coincide with the horizontal and vertical reference directions (X, Y directions) of the apparatus. Correct it. This is performed in order to improve the recognition accuracy in the dimensional information recognition processing in the subsequent steps 3 and 4. Since this automatic skew correction may be performed by some conventionally known methods, detailed description of its contents will be omitted.

Next, the process proceeds to the subroutine of step 3, where the entire image data on which the automatic skew correction is performed is targeted.
Dimension area recognition processing is executed. This process is a process for determining an area (referred to as "dimension area") in which the dimension information exists in the image data of the input construction drawing. The dimension auxiliary line extraction processing in the next step 4 and the dimension position in FIG. This is performed by the recognition unit 6, but the details will be described later.

When the size area recognition process is completed and the size area is determined, the process proceeds to the subroutine of step 4 to execute the size auxiliary line extraction process. This processing is processing for extracting only the auxiliary dimension lines existing therein for each of the dimension areas determined in the processing of step 3 as shown in FIG. 7, and the details thereof will be described later.

When the recognition of the dimension lines of all the dimension areas is completed, the process proceeds to step 5 where the dimension list creating section 7 searches for dimensions and creates dimension lists. That is, as shown in FIG. 8, the dimension line b existing between the pair of dimension extension lines aa is searched. In this search, by scanning within each dimension area recognized in step 3, black dot rows having a predetermined length or more continuous in the direction orthogonal to the dimension extension line are candidates for the dimension line, and both ends thereof are determined. Search for those that are in contact with the extension line.

Then, a dimension list is created by combining the retrieved dimension line and two dimension extension lines corresponding thereto. This dimension list is combined with dimension numbers to be recognized later. This process is repeated until the search for the dimension line is completed, and when the process is completed, the process proceeds to step 7 where the dimension value recognition unit 8 performs dimension value recognition processing. This process is a process of recognizing the dimension numerical value for each dimension line of the dimension list created in step 5 and adding it to the dimension list.
The details will be described later with reference to FIG.

In step 8, the dimensional numerical value of the recognition result may be displayed on the display unit 9 or all dimensional information including the dimensional line and the dimensional auxiliary line may be displayed. And step 9
To determine whether or not to confirm, that is, whether or not the confirmation key of the operation input unit 10 has been pressed (may be an instruction with a mouse cursor), and if it is determined to proceed, proceed to step 10 to confirm the recognition result at that time. Then, it is stored in the definite recognition result storage area of the memory 4 and the processing is terminated.

If the confirm key is not pressed in step 9, it is determined in step 11 whether or not the recognition result is to be corrected, that is, whether the correction input selection key has been pressed (or the mouse cursor may be used), and the correction is made. If judged, the screen for correction input including the image image of the construction drawing input to the display unit 9 in step 13 is displayed. Then, when the operator inputs (including addition, correction and deletion) the correction data (including numerical values) into the memory 4 through the operation input unit 10, the correction data is stored in the memory 4, and the correction result is displayed in the display unit in step 8. Display on 9. The processing of steps 8, 9, 11, 13, 14, 11 is repeated until the correction is completed.

If it is determined in step 11 that the correction is not made, it is determined in step 12 whether or not to reprocess (whether a reprocessing instruction has been issued). If a reprocessing instruction has been issued, go to step 3. Then, the process from the dimension area recognition process to step 10 or 14 is repeated. When the reprocessing instruction is not given in step 12 (the end instruction is given or a predetermined time has passed without giving any instruction), the processing is ended.

Although not shown in this flow chart, the confirmed recognition result is stored in the storage medium of the external storage device 11 in FIG. 1, printed on the paper by the printing device 12, or transmitted from the communication control unit 3. It can also be sent to a host computer or the like.

Next, details of the dimension numerical value recognition processing in step 7 of FIG. 5 will be described with reference to the flowchart of FIG. When this process is started, an area U above the center of the dimension line b is cut out for each dimension line in the dimension list as shown in FIG. 9, and it is checked by scanning whether or not there is a character image. Whether or not there is a character image is determined by whether or not black dots are densely arranged at a predetermined density. If the density is less than the predetermined density, it is determined as noise (dust).

If there is no character image, then the lower region L of the central portion of the dimension line b is cut out as shown in FIG.
Similarly, scan to see if there is an image of the character. If there is no character image there, an error message is displayed on the display unit 9 and the process ends. The content of the error message is "Unable to recognize the numerical value. Check the original image for the numerical value." If there is an actual numerical value but it cannot be recognized, it can be corrected and input in step 14 of FIG.

When there is a character image in either the area U above the dimension line or the area L below the dimension line, as shown in FIG. The width S) is applied to the area S between the pair of dimension extension lines aa by OCR to recognize the character string (character code).

The recognized character string is converted into a numerical value and stored as the numerical value of the corresponding dimension line in the dimension list.
The above-described processing is repeated until the OCR is applied to all the dimension lines S in the dimension list, and when the processing is completed, the process returns to the main routine of FIG.

Next, the dimension area recognition processing of step 3 and the dimension auxiliary line extraction processing of step 4 in the main routine of FIG. 5 will be described. First, the flowchart of the subroutine of the dimension area recognition process is shown in (a) on the left side of FIG. 11, and the explanatory diagrams regarding the process of each step are shown in (b) to (f) on the right side. FIG. 11B shows the image data of the construction drawing that has been input and skew-corrected (construction drawing data).
Is shown in a simplified form.

For this image data, first step 2
In 1, the number of black dots in the horizontal direction (X direction) is measured for each dot line, and the horizontal dot number measurement array data, that is, the horizontal dot histogram of the entire drawing as shown in FIG. create. The number of black dots is measured when the read construction drawing is a positive drawing (a drawing whose brightness is lower than the brightness of the ground), and a negative drawing (a drawing whose brightness is higher than the brightness of the ground). Measure the number of white dots.

Then, in step 22, a portion where the characteristic of the dimension area is seen is searched from the black dot number measurement array.
Then, the range in which this feature can be seen is defined as a provisional size region (horizontal direction) as shown by the dot in (d).

These processes will be further described with reference to FIG. FIG. 8A shows a configuration example of the dimension data, which is composed of a plurality of leader lines a, a plurality of dimension lines b, and dimension characters c. The inner tip of the leader line a may be in contact with the building or separated from it, but there is a plurality of dimension lines b in parallel to each other at a substantially constant distance from the building to the outer tip part side, The line width is thin. This dimension line extends horizontally on the upper or lower side of the building and vertically on the right or left side. The example shown is on the upper side and extends horizontally.

Therefore, the dot histogram which is the horizontal black dot number measurement array of this dimension data is shown in FIG.
(B). That is, a plurality of portions B in which long lines indicating that the number of black dots (the number of black runs) are very large appear in a narrow range in the Y direction appear at substantially constant intervals. This is due to the dimension line b. A is a leader line, and a short line indicating that the number of black dots is extremely small appears in close contact with the Y direction like a vertical thick line. C
Is a portion formed by a dimensional character, and a number of relatively short lines appear adjacent to each other in the Y direction between portions B.

These are the characteristics of the size domain. Therefore,
This dot histogram is scanned, and the range of width W from the outermost long linear portion B to a predetermined width d further preset to the building side than the innermost portion B is provisionally dimensioned. And

Returning to FIG. 11, next, in step 23, vertical black dot number measurement array data, that is, vertical dot histogram of the entire drawing is created. Then, in step 24, the same processing as in the horizontal direction in step 22 is performed to search for a portion in which the characteristic of the dimension area is seen from the black dot number measurement array, and the range is indicated by hatching in (e). To be a temporary dimension area (vertical direction).

Then, in the final step 25, the vertical and horizontal tentative size regions are combined, each region outside the overlapping region is deleted, and halftone dots are applied to FIG. 11 (f). The dimensional area as shown by is recognized and determined. In this example, there are four dimension regions located above, below, to the left and to the right of the building. The coordinate data of the four corners of each size area is stored in the memory. After that, the process returns to the main routine of FIG.

Next, FIG. 13 shows a flowchart of a subroutine of the dimension auxiliary line extraction processing in step 4 of FIG. 5, which will be described with reference to FIGS. 14 to 16. When the dimension auxiliary line extraction processing shown in FIG. 13 is started, first, the coordinate data of the four dimension areas previously recognized in step 31 are read in order.

At step 32, the position of the read dimension area with respect to the building is determined. If the determination result is left or right, the process proceeds to step 33, and horizontal direction black dot number measurement array data is created from the image data in the dimension area. If the determination result is up or down, step 34
Then, the vertical black dot number measurement array data is created from the image data in the dimension area.

For example, as shown in FIG. 14, when the horizontal black dot number measurement array data in the dimension area 21 on the left side of the image 20 of the building in the image data is created, FIG.
A black dot number measurement array (dot histogram) as shown in FIG. In this black dot number measurement array, a narrow peak appears at the position where each dimension extension line is present.

Then, in step 35, the black dot number measuring array is scanned to detect the width of a line in which the number of black dots is equal to or greater than the threshold value and the peak value of the number of black dots in between. The scanning direction is the Y direction (vertical direction) for the horizontal black dot number measurement array and the X direction (horizontal direction) for the vertical black dot number measurement array.

Then, for example, as shown in FIG. 16, scanning is performed in the Y direction with respect to the horizontal black dot number measurement array, and when a line having a black dot number equal to or larger than a threshold value is detected, it is continued in the scanning direction. The detected width D and the peak value of the number of black dots between them (the number of black dots in the second detected line in the illustrated example) are detected and stored.

When the width D is detected, that is, when the line where the number of black dots is equal to or larger than the threshold value is detected and the line cannot be detected first (the line where the number of black dots is smaller than the threshold value is scanned), And when the scanning for the black dot number measurement array in one size area is completed, the process proceeds to step 36.

Then, it is judged whether or not there is a line in which the number of black dots is equal to or larger than the threshold value, and if there is, the process proceeds to step 37 and the width (D shown in FIG. 16) is sufficiently narrow with respect to the height of the peak ( (1/5 or less, 1/10 or less in advance). If it is narrow, it is determined to be a dimension auxiliary line, and its X coordinate (when scanned in the X direction) or Y coordinate (when scanned in the Y direction) is stored in the memory 4 in step 38.
Add to the extension line information list in.

For example, in the case of the left and right dimension areas, FIG.
The Y-direction scanning position of the midpoint P of the width (range in which lines in which the number of black dots in the array is equal to or greater than the threshold value) shown in FIG. In the case of the upper and lower dimension regions, the X-direction scanning position of the similar midpoint of the width D is set as the X coordinate of the dimension extension line.

After that, the process proceeds to step 39 in FIG.
It is judged whether or not the scanning for the black dot number measurement array in one dimension area is completed, and if it is not completed, step 3
Returning to step 5, the subsequent processing is repeated. When the scanning is completed, the process proceeds to step 40, where it is judged whether or not the dimension extension line extraction processing for all the dimension areas has been completed. If not completed, the procedure returns to step 31 to read the coordinate data of the next dimension area, The above process is repeated.

When the dimension auxiliary line extraction processing for all the dimension areas above, below, left and right with respect to the building is completed, the process exits this subroutine and returns to the main routine of FIG. According to this method, it is almost 1 with a simple algorithm.
The coordinates of the point indicated by the 00% dimension line can be recognized. Further, the processing speed is higher than that of the contour line extraction and the thinning processing. Extraction is possible within 1 second even with A3 drawing.

Further, since the dot histogram analysis is performed, it is resistant to noise and blurring, and the coordinate data of the dimension extension line can be acquired accurately. Also, since the memory usage only increases one-dimensionally even if the drawing becomes large,
For example, about 6K bytes for A4 drawings and about 24K for A0 drawings.
Only a part-time job is required.

Here, a concrete example of the above-mentioned dimension area recognition processing and dimension extension line extraction processing for a floor plan of a house which is an actual architectural drawing will be described with reference to FIGS. 17 to 20. FIG. 17 is image image data obtained by reading a floor plan of a house. In order to perform the dimension area recognition processing on this image image data, the number of black dots in the horizontal direction and the number of black dots in the vertical direction are counted, and horizontal black dot number measurement array data shown in (H) of FIG. () Creates vertical black dot number measurement array data (dot histogram).

From the horizontal and vertical black dot number measurement array data, a portion in which the characteristics of the dimension area can be seen is searched for, and the dimension area (dimension information indicated by hatching the image image data in FIG. 18 is indicated. (The range in which exists) is recognized. In this example, there are four dimension regions in the upper, lower, left and right sides of the image of the building.

Since the dimension auxiliary lines are recognized for the image data of each dimension area determined in this way, as shown in FIG. 19, the image data in the dimension areas on the left and right of the building are horizontally aligned. Black dot number measurement array data is created, and vertical black dot number measurement array data is created for each image data in the upper and lower dimension areas of the building.

Then, as shown in FIG. 20 in the vertical direction black dot number measurement array of the one size region, each portion in which a black run in which the number of black dots is the threshold value or more exists in a width sufficiently narrow with respect to the peak. The coordinates of the center point of the
The coordinates (Y coordinate in the case of the horizontal black dot number measurement array) are recognized and stored in the dimension auxiliary line information list.

By the additional function of the dimensional numerical value recognizing unit 8 in FIG. 1, the recognized dimensional numerical values stored in the dimensional list and the position data of the dimensional line and dimensional auxiliary line are used to make other dimensional values in the vicinity. It is also possible to obtain the numerical value of the dimension of the line by calculation. The dimension list is created in the memory 4 for each area obtained by the two dimension extension lines. An example of a method of calculating the dimension value will be described with reference to FIG.

First, the dimension numerical values "1000" and "1500" relating to the dimension lines A and B are recognized, and then they are added to calculate (1000 + 1500 = 2500) as the dimension numerical value relating to the dimension line C. Furthermore, the dimension line C
It is also possible to perform the OCR also on the region corresponding to, and compare it with the calculation result to determine the dimension numerical value related to the dimension line C. After performing all OCR on the area of the dimension line,
When a contradiction occurs after the calculation is performed by this calculation method, the operator can confirm the displayed part and confirm it.

FIG. 22 shows an example of displaying the recognized dimensional numerical values. By displaying the recognizable dimensional numerical values (shown in a frame) as close to the image of the original image (upper, lower, left and right) as possible, comparison can be facilitated. You can

[0073]

As described above, according to the dimensional numerical value recognizing method and recognizing apparatus for drawings of the present invention, the dimensional numerical value which is important information from the image information of the drawing having the dimensional information such as the construction drawing is important. Can be automatically and accurately recognized.

[Brief description of the drawings]

FIG. 1 is a block diagram showing a schematic configuration of an apparatus for implementing a dimension numerical value recognition method of the drawings according to the present invention.

FIG. 2 is a diagram showing a description example of dimensional information in a construction drawing.

FIG. 3 is a diagram showing an example of division of the same size.

FIG. 4 is a diagram showing an example of a construction drawing having dimensional information.

5 is a flow chart of a main routine of a process relating to dimensional numerical value recognition of the drawing by the apparatus shown in FIG.

6 is a flow chart of a subroutine of dimension numerical value recognition processing of step 7 in FIG.

FIG. 7 is an explanatory diagram of dimension auxiliary line extraction.

FIG. 8 is an explanatory diagram of dimension line search.

FIG. 9 is an explanatory diagram of an area for investigating which side of a dimension line has a character image.

FIG. 10 is an explanatory diagram of a region for performing character recognition (OCR).

FIG. 11 is a diagram showing a flow of a subroutine of the dimension area recognition process of step 3 in FIG. 5 and a diagram provided for explaining each step in association with each other.

FIG. 12 is an explanatory diagram for similarly explaining the size area recognition process.

FIG. 13 is a flow chart of a subroutine of dimension auxiliary line extraction processing of step 4 in FIG.

FIG. 14 is a diagram showing an example of a building image and a size region in the input image data of the construction drawing.

FIG. 15 is a diagram showing a horizontal black dot number measurement array in the dimension area of FIG. 9;

FIG. 16 is an enlarged view for explaining a process of recognizing a dimension auxiliary line from the black dot number measurement array.

FIG. 17 is a diagram showing image image data obtained by reading an actual architectural drawing and horizontal and vertical black dot number measurement array data for the image data.

FIG. 18 is a diagram showing an example in which a size area is determined using the black dot number measurement array data in the horizontal and vertical directions of FIG. 17;

FIG. 19 is a diagram in which horizontal or vertical black dot number measurement array data is created for each size area of FIG. 18 based on image data therein.

FIG. 20 is a diagram showing a relationship between a vertical black dot number measurement array in one size region and a threshold value for extracting a size extension line.

FIG. 21 is an explanatory diagram of an example in which a numerical value of a dimension is calculated.

FIG. 22 is a diagram illustrating a display example of recognized dimension numerical values.

[Explanation of symbols]

 1: Overall control unit (CPU) 2: Image reading unit 3: Communication control unit 4: Memory 5: Automatic skew correction unit 6: Dimension position recognition unit 7: Dimension list creation unit 8: Dimension numerical value recognition unit 9: Display unit 10 : Operation input unit 11: External storage device 12: Printing device 13: Bus 20: Image of building 21: Dimension area

────────────────────────────────────────────────── ───

[Procedure amendment]

[Submission date] November 9, 1995

[Procedure amendment 1]

[Document name to be amended] Statement

[Correction target item name] Detailed description of the invention

[Correction method] Change

[Correction contents]

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention recognizes dimensional numerical values in drawings having dimensional information consisting of auxiliary dimension lines, dimensional lines, and numerical values, such as construction drawings such as architectural drawings and various design drawings. The present invention relates to a dimensional numerical value recognition method and apparatus.

[0002]

2. Description of the Related Art In recent years, using CAD systems, etc., construction drawings including houses, buildings, etc., construction drawings including water pipes, gas pipes, piping drawings for power / communication cables, etc. can be easily created, and the data thereof can be obtained. It has been practiced to memorize and store the information when design changes, extension and remodeling are performed. However, the data of the construction drawings are not compatible with each other due to the created system, and cannot be used due to the passage of time or the change of the contractor. In addition, when renovating or remodeling a house, there are many cases where there is only an old architectural drawing drawn on paper, and it has been necessary to redraw all parts including the floor plan of the renovation and remodeling, etc.

Therefore, it has been attempted to read construction drawings such as architectural drawings drawn on paper and recognize and store them as computer processable data, but there is no special method or device for this purpose, A drawing is read by an image scanner, the image data is input to a personal computer or the like, and line segment recognition and pattern recognition are performed using a general figure recognition function. Alternatively, the graphic recognition function can be upgraded by using a high-performance graphic editor to assist, and even if the automatic recognition function is somewhat imperfect, basic vector diagrams such as straight lines and arcs can be converted by raster-to-vector conversion, for example. Some of them can be automatically recognized.

[0004]

However, in such construction drawings and various design drawings, dimensional auxiliary lines,
Many dimensional information consisting of dimensional lines and numerical values is described. This is important information indicating the actual size of each part drawn in the drawing, but the conventional drawing recognition device cannot automatically recognize this size information, especially the size value. Therefore, the operator manually inputs the numerical values one by one from a keyboard or the like.

Even when the dimensional numerical value is read by using the OCR, it is necessary to specify each reading area in which the dimensional numerical value is described. Therefore, when the contents of the drawings having the dimension information are converted into a database that can be processed by a computer, there is a problem that the input of the dimension values is complicated and the input error may occur in the conventional technique.

The present invention has been made in view of the above points, and it is possible to automatically and accurately and quickly recognize a dimensional value which is important information from image information of a drawing having dimensional information such as a construction drawing. The purpose is to do so.

[0007]

In order to achieve the above object, the present invention provides a dimensional numerical value recognizing method and a dimensional numerical value recognizing apparatus in the following drawings. A method of recognizing dimension numerical values of a drawing according to the present invention inputs image information of a drawing having dimension information including a dimension auxiliary line, a dimension line, and a dimension numerical value, extracts each dimension auxiliary line in the input image information, The dimension lines existing between the extracted dimension extension lines are searched, the area on the side where there is a character with respect to the found dimension line is searched, and the dimension is determined near the dimension line on the side where it is determined that there is a character. It is characterized in that character recognition processing is performed on a region between lines to recognize a numerical value.

Further, the drawing numerical value recognizing device according to the present invention includes image information input means for inputting image information of a drawing having size information consisting of auxiliary dimension lines, dimension lines and numerical values, and the inputted image information. Means for extracting each of the dimension auxiliary lines, means for searching the dimension lines existing between the dimension lines extracted by the means, and which side of the searched dimension line the character is in And means for recognizing the dimension numerical value by performing character recognition processing on the area between the dimension extension lines in the vicinity of the dimension line on the side where the character is determined to be present by the means. .

Therefore, it is possible to automatically search the area in which the dimensional numerical value is described and recognize the dimensional numerical value by OCR or the like, so that the dimensional numerical value can be manually input or the reading areas can be individually read. There is no need to specify it.

Further, in the dimensional numerical value recognizing apparatus of this drawing, the data of the dimensional line which can be searched, the data of the two dimensional auxiliary lines corresponding to the dimensional line, and the dimensional numerical value recognized about the dimensional line are put together. It is advisable to provide dimension list creating means for creating the number of dimension lists described above according to the number of dimension lines.
As a result, the numerical values and the dimension lines can be associated with each other on a one-to-one basis.

Further, it is preferable to provide a calculating means for calculating a dimension numerical value related to another dimension line in the vicinity from the already recognized dimension numerical value based on the information of the dimension list created by the dimension list creating means. As a result, it is possible to calculate the dimension values of other dimension lines in the vicinity from the recognized dimension values without performing all the dimension values by character recognition processing such as OCR, thereby reducing the recognition time and the recognition accuracy. Can be improved.

If the image information input means is an image reading means for reading the image of the drawing and inputting the image information of the image, it is possible to easily recognize the dimensional values on the drawing drawn on the paper. Further, if the image information input means is an image information receiving means for receiving and inputting the image information of the drawing by communication, the image information of the drawing is received from another device at a remote place, It is possible to easily recognize the dimensional numerical value of.

Further, by providing a display means for displaying the recognized dimensional numerical value at the same time as the input image information image, it is possible to easily confirm the recognition result and find the erroneous recognition. If the recognition result confirmation means for confirming the recognized dimensional numerical value displayed on the display means by an instruction from the outside is provided, the operator can confirm the recognition result and then confirm it by the instruction.

[0014]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be specifically described below with reference to the drawings. FIG. 1 is a block diagram showing a schematic configuration of a dimensional numerical value recognizing apparatus for carrying out the dimensional numerical value recognizing method according to the present invention, in which a hardware configuration and a software processing function by a microcomputer are shown in a mixed manner.

This apparatus comprises an overall control unit 1, an image reading unit 2, a communication control unit 3, a memory 4, an automatic skew correction unit 5,
It is composed of a dimensional position recognizing unit 6, a dimensional list creating unit 7, a dimensional numerical value recognizing unit 8, a display unit 9, an operation input unit 10, an external storage unit 11, a printing unit 12 and a bus 13 connecting them. In addition, these parts (or devices) and the bus 13
The interface part required between the and is not shown.

The overall control unit 1 is a microcomputer (CPU, RO) for controlling the operation and function of the entire apparatus.
M, RAM, etc., but is abbreviated as "CPU" as a representative). The automatic skew correction section 5, the dimension position recognition section 6, the dimension list creation section 7, and the dimension value recognition section according to the present invention. Each function of 8 can also be realized by the software processing of the CPU.

The image reading section 2 is an image information input means for scanning a drawing having set dimension information, reading the image, and inputting image information. The image reading section 2 includes a scanning optical system and an image sensor such as a CCD. It is a known image scanner including its drive circuit and the like. Also,
It also includes a binarizing circuit that binarizes the read image image information at a predetermined resolution to produce image data of white dots and black dots.

The communication control unit 3 receives the image information from the image reading unit 2 instead of receiving the image information from the image reading unit 2, and receives and inputs the image image information or the encoded image data in which the run length is encoded by communication from the outside. In addition to the means, the dimensional numerical value recognized by this device or dimensional information including it can be transmitted to an external device. Specifically, it includes a FAX modem and personal computer communication control means.

The memory 4 includes image image information read by the image reading unit 2, image image information received by the communication control unit 3 or encoded image data, image data skew-corrected by the automatic skew correction unit 5, and dimensions. A large-capacity RAM or a memory such as a hard disk for storing the dimensional area recognized by the position recognizing unit 6, the information of the dimensional list created by the dimensional list creating means 7, the dimensional numerical value recognized by the dimensional numerical value recognizing unit 8, and the like. is there.

The automatic skew correction unit 5 adjusts the angle of the image data stored in the memory 4 so that the horizontal and vertical line segment directions coincide with the horizontal and vertical reference directions (X and Y directions) of the apparatus. This is for correction, and a known automatic skew correction technique can be used. The image data corrected by the automatic skew correction unit 5 is stored in the memory 4 again.

The dimensional position recognizing unit 6 is a region where the dimensional information in the image data or the encoded image data stored in the memory 4 after the automatic skew correction is present (hereinafter referred to as "dimension region"). )), And a dimension auxiliary line is extracted for each of the dimension regions to recognize the dimension position.

The dimension list creating section 7 includes a dimension position recognizing section 6
The dimension lines between the dimension extension lines extracted by are searched, and the data of the extracted dimension lines, the data of the two dimension extension lines corresponding to the dimension lines, and the dimension values related to the dimension lines (dimensions described later) Dimension list creating means for creating a dimensional list in which the numerical value recognizing unit 8 is recognized.

The dimensional numerical value recognizing unit 8 searches for the area on the side of each dimensional line of the dimensional list created by the dimensional list creating section to find a character, and determines the size of the side on which the character is found. OCR for the area between the dimension lines near the line
Is a dimensional numerical value recognizing means for recognizing a dimensional numerical value by performing character recognition processing according to. Further, it also has a function of a calculating means for calculating the dimension numerical values of other dimension lines in the vicinity from the already recognized dimension numerical values based on the information of the dimension list. Details of these processes of the dimension position recognition unit 6, the dimension list creation unit 7, and the dimension value recognition unit 8 will be described later.

The display unit 9 receives the image data of the drawing which is input from the image reading unit 2 or the communication control unit 3 and is skew-corrected by the automatic skew correction unit 5, the dimension area and the dimension auxiliary recognized by the dimension position recognition unit 6. It is for displaying a line, a dimension line searched by the dimension list creating unit 7, a dimension value recognized by the dimension value recognizing unit 8, and the like, for example, a CRT or a liquid crystal display. In particular, it has a function of displaying the dimensional numerical value recognized by the dimensional numerical value recognizing unit 8 or the dimensional information including the dimensional line and the dimensional auxiliary line simultaneously with the image of the image information of the inputted drawing.

In this case, the recognized dimension information and the dimension information of the image of the inputted image information of the drawing can be displayed adjacent to each other or can be superimposed and displayed.
Then, in order to easily identify the two, it is preferable to display the image image in halftone and display the dimension value of the recognition result in high contrast. In the case where the display unit 9 is a color display device, the distinguishability can be improved by displaying the two images in different colors. Further, when the possibility of erroneous recognition of the recognized dimensional numerical value is high (equal to or more than a predetermined value), it is possible to display that the erroneous recognition is high by blinking the dimensional numerical value.

Further, the display unit 9 also displays a screen for the operator to confirm the recognition result. That is, the dimension value of the recognition result is displayed, and a display such as "Are you sure with this recognition result? (YES / NO)" is displayed. As a result, the operator can confirm whether or not the dimensional numerical value can be accurately recognized.

The operation input unit 10 is for inputting various operation instructions, function selection instructions, edit data, etc., and is a keyboard, a mouse, a touch panel, or the like. The operation input unit 10 also has a function as a display selection unit,
The display state of the display unit 9 can be changed to a display state desired by the operator. For example, by operating a key, it is possible to superimpose and display the input image image of the drawing and the dimension information of the recognition result, or select and display only one of them.

Further, the operation input unit 10 inputs a key or the like for inputting "YES" or "NO" information in response to the above-mentioned "is this recognition result OK? (YES / NO)" display. We also have means. Then, when "YES" is selected, the displayed recognition result is confirmed and the recognition process is ended, and when "NO" is selected, the process is shifted to the re-recognition process or the correction process. As a result, the operator can confirm the content of the recognition result and confirm it.

Further, the operation input unit 10 is provided with a mouse as a pointing device, and the dimension auxiliary line or the dimension line of the input image image of the drawing displayed on the display unit 9 is traced by the mouse cursor. It is also possible to provide a function for inputting position information of the or dimension line.

The external storage device 11 stores the dimensional numerical values recognized by the dimensional numerical value recognizing section 8 and the information of the dimensional list created by the dimensional list creating section 7 into a floppy disk (FD), a magneto-optical disk (OMD) or the like. Is a storage device for storing in a storage medium that can be taken out of the computer. The printing device 12 is a printer or plotter that prints or draws the input image image of the drawing or the recognized dimension information on paper and outputs it.

[0031] Here, a description will be given to the size information of the drawing. Generally, the dimension information is a description indicating the length, size, position, angle, etc. on the drawing. In a construction drawing (mainly an architectural drawing), dimensional information indicating the length is usually represented by a combination of dimensional figures as shown in FIG. That is, a dimension line is drawn between two dimension extension lines that are drawn out from both ends of a portion where the dimension of the target graphic is desired to be designated, and a dimension value is written in the central portion on the dimension line. . That is, the dimensional values and the dimensional lines are in one-to-one correspondence.

FIG. 2A shows an example in which the dimension information is arranged on the upper side of the figure, and FIG. 2B shows an example in which the dimension information is arranged on the lower side of the figure. Although the position of the dimension line with respect to the dimension extension line is different for both, there is a dimension numerical value on the dimension line (on the right side if the dimension line is vertical), and both are closer to the dimension line below the number. . However, in some cases, the dimension number is below the dimension line (left side if the dimension line is vertical).
It may be described in.

FIG. 3 is a display example in the case where a long dimension is divided into a plurality of parts and the individual dimensions are shown.
When the dimensional information shown in is decomposed, it becomes as shown in (b).
FIG. 4 shows an example of a floor plan of an architectural drawing having actual dimension information.

Next, the procedure of the dimensional numerical value recognition process of the construction drawings (mainly architectural drawings of houses, buildings, etc.) by the dimensional numerical value recognizing apparatus shown in FIG. 1 will be described with reference to the flowcharts of FIGS. 5 and 6. To do. In the flowchart of FIG. 5, each step is indicated by “S”. In the example described here, the construction drawing to be recognized is assumed to be a positive drawing.

FIG. 5 shows the main routine of the processing relating to the present invention by the overall control unit (CPU) 1 of the dimension numerical value recognition apparatus for construction drawings according to the present invention. First, step 1
At, the image of the construction drawing set in the image reading unit 2 is read, and the image data obtained by binarizing the image is input and stored in the image data storage area of the memory 4. Alternatively, the communication control unit 3 may externally receive and input image data of the construction drawing (image image data or encoded image data obtained by encoding the run length).

In step 2, the automatic skew correction unit 5 makes the horizontal and vertical line segment directions of the input image data stored in the memory 4 coincide with the horizontal and vertical reference directions (X, Y directions) of the apparatus. Correct it. This is performed in order to improve the recognition accuracy in the dimensional information recognition processing in the subsequent steps 3 and 4. Since this automatic skew correction may be performed by some conventionally known methods, detailed description of its contents will be omitted.

Next, the process proceeds to the subroutine of step 3, where the entire image data on which the automatic skew correction is performed is targeted.
Dimension area recognition processing is executed. This process is a process of determining an area (referred to as a "dimension area") in which the dimension information in the input construction drawing image data exists, and together with the dimension auxiliary line extraction processing in the next step 4, the dimension position in FIG. This is performed by the recognition unit 6, but the details will be described later.

When the size area recognition process is completed and the size area is determined, the process proceeds to the subroutine of step 4 to execute the size auxiliary line extraction process. This processing is processing for extracting only the auxiliary dimension lines existing therein for each of the dimension areas determined in the processing of step 3 as shown in FIG. 7, and the details thereof will be described later.

When the recognition of the dimension lines of all the dimension areas is completed, the process proceeds to step 5 where the dimension list creating section 7 searches for dimensions and creates dimension lists. That is, as shown in FIG. 8, the dimension line b existing between the pair of dimension extension lines aa is searched. In this search, by scanning within each dimension area recognized in step 3, black dot rows having a predetermined length or more continuous in the direction orthogonal to the dimension extension line are candidates for the dimension line, and both ends thereof are determined. Search for those that are in contact with the extension line.

Then, a dimension list is created by combining the retrieved dimension line and two dimension extension lines corresponding thereto. This dimension list is combined with dimension numbers to be recognized later. This process is repeated until the search for the dimension line is completed, and when the process is completed, the process proceeds to step 7 where the dimension value recognition unit 8 performs dimension value recognition processing. This process is a process of recognizing the dimension numerical value for each dimension line of the dimension list created in step 5 and adding it to the dimension list.
The details will be described later with reference to FIG.

In step 8, the dimensional numerical value of the recognition result may be displayed on the display unit 9 or all dimensional information including the dimensional line and the dimensional auxiliary line may be displayed. And step 9
To determine whether or not to confirm, that is, whether or not the confirmation key of the operation input unit 10 has been pressed (may be an instruction with a mouse cursor), and if it is determined to proceed, proceed to step 10 to confirm the recognition result at that time. Then, it is stored in the definite recognition result storage area of the memory 4 and the processing is terminated.

If the confirm key is not pressed in step 9, it is determined in step 11 whether or not the recognition result is to be corrected, that is, whether the correction input selection key has been pressed (or the mouse cursor may be used), and the correction is made. If judged, the screen for correction input including the image image of the construction drawing input to the display unit 9 in step 13 is displayed. Then, when the operator inputs (including addition, correction and deletion) the correction data (including numerical values) into the memory 4 through the operation input unit 10, the correction data is stored in the memory 4, and the correction result is displayed in the display unit in step 8. Display on 9. The processing of steps 8, 9, 11, 13, 14, 11 is repeated until the correction is completed.

If it is determined in step 11 that the correction is not made, it is determined in step 12 whether or not to reprocess (whether a reprocessing instruction has been issued). If a reprocessing instruction has been issued, go to step 3. Then, the process from the dimension area recognition process to step 10 or 14 is repeated. When the reprocessing instruction is not given in step 12 (the end instruction is given or a predetermined time has passed without giving any instruction), the processing is ended.

Although not shown in this flow chart, the confirmed recognition result is stored in the storage medium of the external storage device 11 in FIG. 1, printed on the paper by the printing device 12, or transmitted from the communication control unit 3. It can also be sent to a host computer or the like.

Next, details of the dimension numerical value recognition processing in step 7 of FIG. 5 will be described with reference to the flowchart of FIG. When this process is started, an area U above the center of the dimension line b is cut out for each dimension line in the dimension list as shown in FIG. 9, and it is checked by scanning whether or not there is a character image. Whether or not there is a character image is determined by whether or not black dots are densely arranged at a predetermined density. If the density is less than the predetermined density, it is determined as noise (dust).

If there is no character image, then the lower region L of the central portion of the dimension line b is cut out as shown in FIG.
Similarly, scan to see if there is an image of the character. If there is no character image there, an error message is displayed on the display unit 9 and the process ends. The content of the error message is "Unable to recognize the numerical value. Check the original image for the numerical value." If there is an actual numerical value but it cannot be recognized , it can be corrected and input in step 14 of FIG.

When there is a character image in either the area U above the dimension line or the area L below the dimension line, as shown in FIG. The width S) is applied to the area S between the pair of dimension extension lines aa by OCR to recognize the character string (character code).

The recognized character string is converted into a numerical value and stored as the numerical value of the corresponding dimension line in the dimension list.
The above-described processing is repeated until the OCR is applied to all the dimension lines S in the dimension list, and when the processing is completed, the process returns to the main routine of FIG.

Next, the dimension area recognition processing of step 3 and the dimension auxiliary line extraction processing of step 4 in the main routine of FIG. 5 will be described. First, a flow chart of a subroutine of the dimension area recognition processing is shown in (a) on the left side of FIG. 11, and explanatory diagrams regarding the processing of each step are shown in (b) to (f) on the right side. FIG. 11B shows the image data of the construction drawing that has been input and skew-corrected (construction drawing data).
Is shown in a simplified form.

For this image data, first step 2
In 1, the number of black dots in the horizontal direction (X direction) is measured for each dot line, and the horizontal dot number measurement array data, that is, the horizontal dot histogram of the entire drawing as shown in FIG. create. The number of black dots is measured when the read construction drawing is a positive drawing (a drawing whose brightness is lower than the brightness of the ground), and a negative drawing (a drawing whose brightness is higher than the brightness of the ground). Measure the number of white dots.

Then, in step 22, a portion in which the characteristic of the dimension area is seen is searched from the black dot number measurement array.
Then, the range in which this feature can be seen is defined as a provisional size region (horizontal direction) as shown by the dot in (d).

These processes will be further described with reference to FIG. FIG. 12A shows a configuration example of the dimension data, which is composed of a plurality of leader lines a, a plurality of dimension lines b, and dimension characters c. The inner tip of the leader line a may be in contact with the building or separated from it, but there is a plurality of dimension lines b in parallel to each other at a substantially constant distance from the building to the outer tip part side, The line width is thin. This dimension line extends horizontally on the upper or lower side of the building and vertically on the right or left side. The example shown is on the upper side and extends horizontally.

Therefore, the dot histogram which is the horizontal black dot number measurement array of this dimension data is shown in FIG.
(B). That is, a plurality of portions B in which long lines indicating that the number of black dots (the number of black runs) are very large appear in a narrow range in the Y direction appear at substantially constant intervals. This is due to the dimension line b. A is a leader line, and a short line indicating that the number of black dots is extremely small appears in close contact with the Y direction like a vertical thick line. C
Is a portion formed by a dimensional character, and a number of relatively short lines appear adjacent to each other in the Y direction between portions B.

These are the characteristics of the size domain. Therefore,
This dot histogram is scanned, and the range of width W from the outermost long linear portion B to a predetermined width d further preset to the building side than the innermost portion B is provisionally dimensioned. And

Returning to FIG. 11, next, in step 23, vertical black dot number measurement array data, that is, vertical dot histogram of the entire drawing is created. Then, in step 24, the same processing as in the horizontal direction in step 22 is performed to search for a portion in which the characteristic of the dimension area is seen from the black dot number measurement array, and the range is indicated by hatching in (e). To be a temporary dimension area (vertical direction).

Then, in the final step 25, the vertical and horizontal tentative size regions are combined, each region outside the overlapping region is deleted, and halftone dots are applied to FIG. 11 (f). The dimensional area as shown by is recognized and determined. In this example, there are four dimension regions located above, below, to the left and to the right of the building. The coordinate data of the four corners of each size area is stored in the memory. After that, the process returns to the main routine of FIG.

Next, FIG. 13 shows a flowchart of a subroutine of the dimension auxiliary line extraction processing in step 4 of FIG. 5, which will be described with reference to FIGS. 14 to 16. When the dimension auxiliary line extraction processing shown in FIG. 13 is started, first, the coordinate data of the four dimension areas previously recognized in step 31 are read in order.

In step 32, the position of the read dimension area with respect to the building is determined. If the determination result is left or right, the process proceeds to step 33, and horizontal direction black dot number measurement array data is created from the image data in the dimension area. If the determination result is up or down, step 34
Then, the vertical black dot number measurement array data is created from the image data in the dimension area.

For example, as shown in FIG. 14, when the horizontal black dot number measurement array data in the dimension area 21 on the left side of the image 20 of the building in the image data is created, FIG.
A black dot number measurement array (dot histogram) as shown in FIG. In this black dot number measurement array, a narrow peak appears at the position where each dimension extension line is present.

Therefore, in step 35, the black dot number measuring array is scanned to detect the width of a line in which the number of black dots is greater than or equal to the threshold value and the peak value of the number of black dots between them. The scanning direction is the Y direction (vertical direction) for the horizontal black dot number measurement array and the X direction (horizontal direction) for the vertical black dot number measurement array.

Then, for example, as shown in FIG. 16, scanning is performed in the Y direction with respect to the horizontal black dot number measurement array, and when a line having a black dot number equal to or larger than a threshold value is detected, it is continued in the scanning direction. The detected width D and the peak value of the number of black dots between them (the number of black dots in the second detected line in the illustrated example) are detected and stored.

When the width D is detected, that is, when the line where the number of black dots is equal to or larger than the threshold value is detected and the line cannot be detected first (the line where the number of black dots is smaller than the threshold value is scanned), And when the scanning for the black dot number measurement array in one size area is completed, the process proceeds to step 36.

Then, it is judged whether or not there is a line in which the number of black dots is equal to or larger than the threshold value, and if there is, the process proceeds to step 37 and the width (D shown in FIG. 16) is sufficiently narrow with respect to the height of the peak ( (1/5 or less, 1/10 or less in advance). If it is narrow, it is determined to be a dimension auxiliary line, and its X coordinate (when scanned in the X direction) or Y coordinate (when scanned in the Y direction) is stored in the memory 4 in step 38.
Add to the extension line information list in.

For example, in the case of the left and right dimension areas, FIG.
The Y-direction scanning position of the midpoint P of the width (range in which lines in which the number of black dots in the array is equal to or greater than the threshold value) shown in FIG. In the case of the upper and lower dimension regions, the X-direction scanning position of the similar midpoint of the width D is set as the X coordinate of the dimension extension line.

After that, the process proceeds to step 39 in FIG.
It is judged whether or not the scanning for the black dot number measurement array in one dimension area is completed, and if it is not completed, step 3
Returning to step 5, the subsequent processing is repeated. When the scan ends
In step 40, it is determined whether or not the dimension extension line extraction processing has been completed for all dimension areas. If not completed, the procedure returns to step 31, the coordinate data of the next dimension area is read, and the above-described processing is performed. repeat.

When the dimension auxiliary line extraction processing for all the dimension areas above, below, left and right with respect to the building is completed, the process exits this subroutine and returns to the main routine of FIG. According to this method, it is almost 1 with a simple algorithm.
The coordinates of the point indicated by the 00% dimension line can be recognized. Further, the processing speed is higher than that of the contour line extraction and the thinning processing. Extraction is possible within 1 second even with A3 drawing.

Further, since the dot histogram analysis is performed, it is resistant to noise and blurring, and the coordinate data of the dimension extension line can be acquired accurately. Also, since the memory usage only increases one-dimensionally even if the drawing becomes large,
For example, about 6K bytes for A4 drawings and about 24K for A0 drawings.
Only a part-time job is required.

Here, a concrete example of the above-mentioned dimension area recognition processing and dimension extension line extraction processing for a floor plan of a house which is an actual architectural drawing will be described with reference to FIGS. 17 to 20. FIG. 17 is image image data obtained by reading a floor plan of a house. In order to perform the dimension area recognition processing on this image image data, the number of black dots in the horizontal direction and the number of black dots in the vertical direction are counted, and horizontal black dot number measurement array data shown in (H) of FIG. () Creates vertical black dot number measurement array data (dot histogram).

From the horizontal and vertical black dot number measurement array data, a portion in which the characteristics of the dimension area can be seen is searched for, and the dimension area (dimension information indicated by hatching the image image data in FIG. 18 is indicated. (The range in which exists) is recognized. In this example, there are four dimension regions in the upper, lower, left and right sides of the image of the building.

Since the dimension auxiliary lines are recognized for the image data of each dimension area determined in this way, as shown in FIG. 19, the image data in the dimension areas on the left and right of the building are horizontally aligned. Black dot number measurement array data is created, and vertical black dot number measurement array data is created for each image data in the upper and lower dimension areas of the building.

[0071] Then, as shown in a vertical black dots counting sequence of the one dimension region in FIG. 20, the portion where the number of black dots present in sufficient narrower than the black run thresholds the peak The coordinates of the center point of the
The coordinates (Y coordinate in the case of the horizontal black dot number measurement array) are recognized and stored in the dimension auxiliary line information list.

By the additional function of the dimensional numerical value recognizing unit 8 in FIG . 1, the recognized dimensional numerical values stored in the dimensional list and the other dimensional values in the vicinity are utilized by utilizing the position data of the dimensional lines and dimensional auxiliary lines. the dimension numbers according to the line can also result determined <br/> Mel that operation. The dimension list is created in the memory 4 for each area obtained by the two dimension extension lines. An example of a method of calculating the dimension value will be described with reference to FIG.

[0073] First of all, the dimension values related to dimension lines A and B
Recognize "1000" and "1500", then
Add and measure (1000 + 1500 = 2500)
It is calculated as the dimensional numerical value related to the line C. Furthermore, the dimension line C
After performing OCR in the area for, compare with the calculation result
Then, even if the dimensional numerical value related to the dimension line C is determined
Good. OCR for all dimension areasDone
Later, if there is a contradiction when the calculation is performed using this calculation method,
Display the part of, and let the operator confirm
Can also.

FIG . 22 shows a display example of the recognized dimensional numerical values. The recognized dimensional numerical values (enclosed in a frame) are displayed as close to the image of the original image (upper, lower, left and right) as possible to facilitate comparison. You can

[0075]

As described above, according to the dimensional numerical value recognizing method and recognizing apparatus for drawings of the present invention, the dimensional numerical value which is important information from the image information of the drawing having the dimensional information such as the construction drawing is important. Can be automatically and accurately recognized.

[Procedure amendment 2]

[Document name to be amended] Statement

[Correction target item name] Fig. 4

[Correction method] Change

[Correction contents]

FIG. 4 is a diagram showing an example of a construction drawing having dimensional information.

[Procedure 3]

[Document name to be amended] Drawing

[Correction target item name] Figure 2

[Correction method] Change

[Correction contents]

[Fig. 2]

[Procedure amendment 4]

[Document name to be amended] Drawing

[Correction target item name] Figure 3

[Correction method] Change

[Correction contents]

[Figure 3]

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Internal reference number FI Technical display location G06F 15/60 606F 680B

Claims (8)

[Claims] 1. Inputting image information of a drawing having dimension information including dimension extension lines, dimension lines, and dimension numerical values, extracting each dimension extension line in the input image information, and between the extracted dimension extension lines. Search for the dimension line existing in, and search for the area on the side where there is a character with respect to the dimension line that can be searched, and for the area between the dimension lines near the dimension line on the side where it is determined that there is a character A method for recognizing dimensional numerical values in a drawing, characterized in that dimensional numerical values are recognized by performing character recognition processing. 2. Image information input means for inputting image information of a drawing having dimension information consisting of dimension extension lines, dimension lines, and dimension numerical values, and means for extracting each dimension extension line from the input image information. A means for searching for a dimension line existing between the dimension lines extracted by the means, a means for searching which side of the dimension line can be searched for the character, and the means for character A dimensional numerical value recognizing device for a drawing, comprising means for recognizing a dimensional numerical value by performing character recognition processing on a region between auxiliary dimension lines in the vicinity of a dimensional line judged to be present. 3. The dimension numerical value recognizing device for a drawing according to claim 2, wherein the data of the dimension line that can be searched, the data of two dimension auxiliary lines corresponding to the dimension line, and the dimension line are recognized. A dimension numerical value recognizing device for a drawing, which is provided with a dimension list creating means for creating a dimension list including the dimension numerical values as many as the number of dimension lines. 4. A dimension numerical value recognizing device for a drawing according to claim 3, wherein the dimension numerical value already recognized is related to other dimension lines in the vicinity based on the information of the dimension list created by the dimension list creating means. A dimensional numerical value recognizing means provided with a calculating means for calculating a dimensional numerical value. 5. The dimensional numerical value recognition device for a drawing according to claim 2, wherein the image information input means is an image reading means for reading an image of the drawing and inputting the image information of the image. . 6. The dimensional numerical value recognizing device for a drawing according to claim 2, wherein the image information input means is an image information receiving means for receiving and inputting image information of the drawing by communication. . 7. The dimensional numerical value recognizing device according to claim 1, further comprising display means for displaying the recognized dimensional numerical value at the same time as the image of the input image information. A numerical value recognition device for drawings. 8. The dimension numerical value recognizing device according to claim 7, further comprising a recognition result confirmation means for confirming the recognized dimension numerical value displayed on the display means by an external instruction. Dimensional number recognition device.