JPH02293980A - Image processor - Google Patents

Abstract

PURPOSE:To improve the operability by providing an input part for detecting a position of a designated point of a designated area on an image pattern and using an image processing part as an image processing part for detecting the processing information, based on the position of the designated point. CONSTITUTION:This processor is provided with an input part 56 for detecting a position of a designated point of a designated area on an image pattern, and an image processing part 54 is used as an image processing part for detecting the processing information, based on at least the position of the designated point. The input part 56 detects the position of the designated point of the designated area, and the image processing part 54 detects the processing information of the designated area designated by this designated point, based on the position of the designated point detected. In such a way, since the image processing part 54 detects the processing information of the designated area, it will suffice that an operator only executes a work for indicating the designated point, and it is unnecessary to input the processing information of the designated area.

Description

DETAILED DESCRIPTION OF THE INVENTION (Industrial Field of Application) The present invention applies to image processing devices, such as document file systems and character recognition devices.

(Prior Art) A number of systems have been proposed in the past that perform image processing on input image patterns according to instructions from an operator. Conventional systems of this type include, for example, document file syntax and character recognition devices as explained below. First, the conventional document file system will be explained.

Figure 11 shows an example of a form used in conventional document file purchasing. When filing an image pattern of a form 10 that includes a mixture of text and photos as shown in the same figure, the text portion 12
It is desirable to obtain a pattern that has undergone simple value conversion for the image pattern of , and a pattern that has been binarized by dithering for the image pattern of the photo portion 14. It is necessary to perform different binarization processing. However, a simple and general-purpose image determination technique for determining portions 12 and 14 has not yet been established. For this reason, in the conventional device, an image pattern of an electrical signal obtained by photoelectrically converting the optical information from the form 10 is displayed on the display screen, and the operator can select the area of the text portion 12 and the photograph portion 14 on the display screen. At the same time, we also specified the processing method for each part 12J4. The first step is to specify these areas and processing method information.
This will be explained with reference to Figure 2. FIG. 12 is a diagram showing an example of a display screen in a conventional file device. As shown in the figure, the display screen 16 displays a form pattern 18 that is an image pattern of the form 10, a processing instruction area 20a120b for specifying a processing method, and a cursor 22 for specifying a point on the display screen. are doing. 2
4 and 26 indicate a character part pattern and a photograph part pattern which are image patterns of the character part 12 and the photograph part 14. The operator moves the cursor 22 by operating a mouse (not shown) while looking at the display screen 16. and cursor 2
2 into the processing instruction area 20a, and click/link the mouse button while pointing within the area 20a with the cursor 22. Furthermore, move the cursor 22 to the first designated point 30 in the character area 28, and hold down the mouse button while pointing the designated point 30 with the cursor 22. Next, while holding down the mouse button, move the cursor 22 to the second designated point 32 in the character area 28, and release the mouse button while pointing the designated point 32 with the cursor 22. This series of operations completes the task of specifying designated points 30 and 32 in the character area 28 and the task of specifying the processing method within the character area i'5+28. Similarly, the operator clicks the mouse button while pointing within the area 20b with the cursor 22. This click specifies the processing method within the photo area 34. Next, move the carnaire 22 to the first designated point 36 in the photo area 34, and hold down the mouse button while pointing to the designated point 36. Then, while holding down the mouse button, move the cursor 22 to the second specified point 38 in the photo area 34 and release the mouse button. In the conventional file device, the form pattern in the area 28 is simply divided into two based on the instruction information in the areas 28 and 34 and the processing instruction information in the areas 28 and 34 output from the mouse by the operator's work as described above. The form patterns 18 in the area 34 are converted into values by dithering, and each of the binarized patterns is combined, encoded, and stored in a magnetic file. Next, a conventional character recognition system will be explained.

In Japanese language recognition processing, it is very important to provide information on whether a document is written horizontally or vertically to the reader before starting the processing. However, a general-purpose technique that allows the device itself to analyze a document image and determine whether the document is written horizontally or vertically has not yet been developed. Therefore, a method is generally adopted in which the operator specifies the character arrangement, whether it is horizontal or vertical writing. The following describes the recognition processing area provided in a conventional character recognition device and the process of instructing the processing method of the recognition processing area with reference to the drawings. Figures 13(A) and 13(B) show examples of forms used in conventional character recognition devices. In FIGS. 13(A) and 13(B), 40 and 42 indicate forms for horizontally written and vertically written documents. Furthermore, FIGS. 14(A) and 14(B) show examples of display screens in conventional character recognition devices. Figure 14 (A) and (
8) 44 indicates the display screen of the character recognition device. On the display screen 44, the outline of the form to be recognized, process designation areas 46a and 46b, and a cursor 22 are displayed. 48 in Figure (A) indicates the pattern of the form 40, and 50 in Figure (B) indicates the pattern of the form 42. Even in conventional character recognition devices, an image pattern of an electrical signal obtained by photoelectrically converting optical information from a form, such as pattern 48, is used.
Display screen 44 {This is displayed. Then, the operator specifies an area for the character portion 12 on the display screen 44 and also specifies a processing method for the character portion 12 according to the character arrangement. The operator operates a mouse (not shown) while looking at the display screen 44, and moves the cursor 22 to within the processing instruction area, or within the area 46a in the case of a horizontally written document. Then, while pointing within the processing instruction area with the cursor 22, click the mouse button. Next, move the cursor 22 and press down the mouse button with your finger while pointing the cursor 22 to the first designated point 30. Roughly move the carnage 22 while holding down the mouse button, and release your finger from the mouse button while pointing to the second designated point 30. Through this series of operations, the operations of specifying the specified points 30 and 32 and the operation of specifying the processing method within the character area 28 are completed. In the conventional character recognition device, the instruction information in the area 28 and the area V output from the mouse by the operator's work as described above are used.
Based on the processing instruction information of i28, recognition processing was performed by cutting out patterns (cutting patterns) for each unit of recognition target from patterns within two days, for example, 48. (Problem to be Solved by the Invention) However, in all of the conventional image processing devices described above, the cursor is moved to the area specification point and then moved to the processing method specification area in order to specify the processing method for the specified area. This increases the amount of mouse movement required. Moreover, since you have to operate the mouse button to specify the area specification point and the processing method,
The number of mouse button operations increases. For this reason, the task of instructing these designated points and processing methods is complicated for the operator, which can easily lead to operator errors, and the processing efficiency of these instruction tasks is also low. SUMMARY OF THE INVENTION An object of the present invention is to provide an image processing apparatus that determines a processing method based on the position of an area designation point in order to solve the above-mentioned conventional problems. (Means for Solving the Problems) In order to achieve this object, the image processing device of the present invention includes a preprocessing unit for obtaining an image pattern of an information medium, and an image processing unit for performing image processing within a designated area of the image pattern. An image processing device comprising an image processing section that performs processing according to information, an input section for detecting a specified point in a specified area on an image pattern; It is characterized by an image processing unit that detects processing information based on In carrying out the present invention, the preprocessing section is a preprocessing section that obtains a multivalued image pattern, and the image processing section is a first binarization section that performs simple binarization processing of the image pattern. a second binarization unit that performs binarization processing using dithering of the image pattern; and an image synthesis unit that synthesizes the image pattern that has been subjected to simple digitization processing and the image pattern that has been subjected to binarization processing that has been performed using dithering. and a control unit that detects processing information and encodes the synthesized image pattern;
It is preferable to consist of a storage unit that stores encoded image patterns. In carrying out the present invention, the preprocessing section is a preprocessing section that obtains a binary image pattern, and the image processing section is a recognition section that cuts out a recognition target pattern from the image pattern and performs recognition processing on the recognition target pattern. It is preferable that the system be comprised of a control section that detects information. (Operation) According to the image processing device configured as described above, the input section detects the position M of the designated point in the designated area, and the image processing section performs the designation by the designated point based on the position of the detected designated point. Detects the processing information of the specified area that is lost. Therefore, since the image processing unit detects the processing information of the specified area, the operator only needs to specify the specified point and there is no need to input the processing information of the specified area. (Examples) Examples of the present invention will be described below with reference to the drawings. Note that the drawings are merely shown schematically to the extent that the present invention can be understood, and therefore the configuration, operation, flow of operation, input/output signals, and numerical conditions of each component are not limited to the illustrated examples. ．． Second Example This embodiment describes an example in which this invention is applied to a document file device. FIG. 1 is a functional block diagram schematically showing the configuration of a first embodiment of the present invention. As shown in FIG. 1, the blind device of this embodiment includes a preprocessing unit 52 for obtaining an image pattern of an information medium, and an image processing unit 52 for performing image processing within a specified area of the image pattern in accordance with processing information. It further has an input section 56 for detecting the position of a specified point in a specified area on the image pattern, and the image processing section 54 is specified by the specified point based on the position of the specified point. It has an image processing unit that outputs processing information for a specified area. In this embodiment, the preprocessing unit 52 is a preprocessing unit that obtains a multivalued image pattern, and the photoelectric conversion unit 58 and the image memory 6
Construct from 0. Roughly speaking, the image processing section 54 includes a first binarization section 62 that performs simple binarization processing of an image pattern, a second binarization section 64 that performs a binarization processing of the image pattern by dithering, and a simple binarization section 64 that performs a simple binarization processing of the image pattern. An image synthesis unit 66 that synthesizes an image pattern that has been subjected to a binarization process and an image pattern that has been subjected to a binarization process using dither, and a control unit that detects processing information and encodes the synthesized image pattern. 68, and a storage section 70 for storing encoded image patterns. The input section 56 includes a display section 72 that displays a cursor and image pattern for designating a specified point, and an operation section 74 that controls the display position of the cursor and detects the position M of the specified point. The operating unit 74 includes a mouse (not shown), and by operating the mouse, the operator instructs the cursor display position M18 to be moved to the specified point a.
You may also use a keyboard instead of a mouse. The operation of this embodiment will be explained below. In this example, the specified area is a rectangular area whose diagonal is the line segment connecting the first and second specified points, and when the direction of the vector whose starting and ending points are the first and second specified points is right, the first It is assumed that a rule has been predetermined to simply binarize the specified area specified by the second specified point and to binarize the specified area by dithering when the direction of the vector is leftward, and the operator follows this rule. The equipment of this example is operated for 1ft. The operator places the form 10 (first
(See Figure 1). And Oberake is the operation part 7
A reading start signal is input to the device of the embodiment via the mouse in step 4. Upon input of the reading start signal, the photoelectric conversion unit 58 starts optical scanning of the form. The optical information L from the form is then converted into a multivalued electrical signal, and an image pattern consisting of this electrical signal is output. Image memory 60 stores image patterns from photoelectric conversion section 58.

In this embodiment, the image pattern of one pixel is made into a 4-bit electrical signal, and the density change of one pixel is expressed in 16 gradations. The second binarization unit 64 binarizes the multivalued image pattern in the image memory 60 by dithering and transfers the dithered binarized image pattern to the display unit 72. The display unit 72 displays the image pattern of the entire document based on the transferred image pattern. FIG. 2 is a diagram showing an example of a display screen in this embodiment, in which the form pattern 18 is displayed on the display screen 76 of the display unit 72.
A state in which a cursor 22 and a cursor 22 are displayed is shown.

When the operator finds a character part on the display screen 76, in order to designate the designated area 28 for this character part, the operator first displays the cursor 22 to the 100th place of the first specified point 30, which corresponds to the upper left corner point of the specified area 28. Move position M. When the operator operates the mouse on the operation section 74, the mouse outputs a control signal for controlling the cursor display position to the display section 72 in response to the operation.
The display section 72 displays a cursor 22 at a display position corresponding to a control signal from the mouse.

When the position of the arrow tip of the cursor 22 reaches 100, which indicates the first designated point 30, the operator presses down the mouse button (turns the mouse button on) in order to output a designated point designation signal. When the mouse detects this on state, it outputs the first specified point instruction signal. T! When the designated point designation signal is input, the processing unit 74 detects the arrow tip position M of the cursor 22 at this time as the position of the designated point 30 and outputs it to the control unit 68. Further, in order to output an instruction signal for the second specified point 32 corresponding to the lower right corner point of the specified area 28 of the character part, the operator moves the cursor 22 by moving the mouse Vt while holding down the mouse button, and moves the cursor 22 to the arrow tip of the cursor 22. When the position reaches the second specified point 32 in the specified area 28, press your finger %H from the mouse button (turn the mouse button off). The mouse recognizes this off state and outputs a second specified point instruction signal. When the designated point instruction signal is input, the operating section 74 detects the current arrow tip position of the cursor 22 as M% of the second designated point 32, and outputs it to the control section 68. In the above example, the mouse outputs the first and second specified point instruction signals when it detects the on and off states, but when it detects the off and on states or when it detects a mouse button click, The method of outputting these instruction signals may be changed as desired, such as outputting the first and second designated point instruction signals. After the operator finishes specifying the specified area 28 for the text portion, he next finds the photo area on the display screen 76, and in order to specify the specified area 36 for the photo area, he first selects the first point corresponding to the upper right corner point of the specified area 36. Move the arrow tip M of the cursor 22 to the designated point 36. Then, in order to output a designated point designation signal, the mouse button is turned on while pointing the first designated point 36 at the arrow tip position of the cursor 22.
In order to output the instruction signal of the second specified point 32 corresponding to the lower left corner point of the specified area 34 of the photograph part, the rough overlay moves the tip of the arrow of the cursor 22 to the second specified point 38 while keeping the mouse button on. move it to the indicated position,
Turn off the mouse button while pointing the second designated point 38 at the arrow tip position of the cursor 22. In this way, the digit M of the specified point in each specified area is input to the control section 68 via the input section 56. The control unit 68 sequentially inputs the digit M of the specified point from the operation unit 74 and stores it. When the positions of a predetermined number of specified points, for example, the first and second specified points 30 and 32 for specifying the specified area are input, detection of processing information for the specified area on the second day is started. The control unit 68 can detect processing information in parallel with inputting the designated point position from the input unit 56. Figure 3 is a diagram showing the flow of processing information detection operations in the control unit. When the control unit 68 inputs a predetermined number of designated points necessary to designate one designated area, for example, the positions of two designated points 30 and 32, the controller 68 controls the designated area designated by the predetermined number of designated points. Start detecting processing information to close. still,
The image memory 60 and the display screen 76 each have X-Y
It is assumed that a coordinate system is set virtually, and images of 100 corresponding pixels are stored and displayed at the same X, Y coordinates. The control unit 68 that has started the detection first sets the first and second specified points as the starting point and V! Detect the direction of the point vector (S
l). If the digits M of the first and second designated points are (X, y1) and (X2,'/z), then the direction of the vector is! )
When Xz>X+, turn right and i1) X 2 < X
When it is +, it faces left. Next, the control unit 68 selects processing information predetermined in association with the direction of the detected vector (S2). In this embodiment, if the direction of the vector is rightward or leftward, processing information indicating that simple binarization processing or dithering binarization processing is to be performed on the specified region is selected. Then, the selected processing information is saved together with the positions of the first and second specified points, and processing information detection regarding the specified areas of the first and second specified points is completed. The operator uses the display screen 46
After specifying all the specified points in the specified areas of the upper text and photo areas, click the mouse button twice to input a signal indicating completion of the specified points to the control unit 68. When the mouse detects two clicks, it outputs an instruction end signal. When the control section 68 finishes processing all the processing information regarding the input specified point and receives an instruction end signal, it outputs the position of the specified point and the processing information to the image composition section 66. The image synthesizing section 66 outputs the digit M of the specified point of the specified area indicating that simple binarization processing is to be performed to the first binarization section 62. The first binarization unit 62 reads from the image memory 60 a multi-valued image pattern within a specified area specified by the position of the specified point input from the image synthesis unit 66, and converts it into simple binarization to create a simple binary image pattern. save. On the other hand, the second binarization unit 64 stores the dithered binary image pattern when transferred to the display unit 72. When the first binarization unit 62 finishes the simple binarization process within the specified area, the image synthesis unit 66 generates a specified area having processing information for performing the simple binarization process. reads a binary image pattern of a designated area having processing information for performing a dithering binarization process from the first binarization unit 64, and synthesizes these binary image patterns.

Then, the synthesized binary image pattern is encoded and the encoded pattern is stored in the storage unit 70, for example, a magnetic disk.

The above-described flow of operations for processing a single form is repeated to process a plurality of forms.

According to this embodiment, it is possible to omit the work of inputting processing information (whether it is a character processing area that performs simple binarization or a photo processing area that performs binarization using dithering) that was performed in conventional devices. Therefore, it is possible to improve the work efficiency of the operator and provide a device that is easy to operate. In addition, in conventional devices, areas for instructing processing information (character processing instruction area and photo processing instruction area) were displayed on the display screen, but in this embodiment, these instruction areas can be omitted, thus reducing the size of the display. This allows the production cost of the device to be reduced. No.! FIG. 4 is a functional block diagram schematically showing the configuration of a second embodiment of the present invention. Note that constituent components corresponding to those of the first embodiment are indicated with the same reference numerals. The blind device of the second embodiment has the same configuration as the first embodiment except that the input section 77 is composed of an input device called a tablet and a keyboard, but the operation is slightly different from that of the first embodiment.
Below, we will mainly explain the points that are different from the first embodiment, and will omit a detailed explanation of the points that are similar to the first embodiment. FIG. 5 is a diagram showing a state in which a form is placed on the tablet of the input section of the second embodiment, and shows the tablet 7 of the input section 77.
7a shows a state in which the form 10 is fixed at a predetermined position.
Is form 10 a tablet? ? Tablet via the fixture 78 provided in a? Fixed to 7a. A coordinate system corresponding to the X-Y coordinate system set on the image memory 60 is set on the tablet 77a, and a pixel position on the image memory is associated with a pixel position on the form 10 in a one-to-one correspondence. There is. The operation of the second embodiment will be explained with reference to FIG. The operator fixes the form 10 in a predetermined position on the tablet 77a. Then, a first specified point 30 corresponding to the upper left corner point of the specified area of the character portion 12 of the form 10 is pressed using a pencil-like instrument (not shown). Tablet? ? a detects this pressing position M and outputs it as the position of the first specified point 30 of the suppressing position. Next, the operator selects a second specified point 32 corresponding to the lower right corner point of the specified area of the character portion 12 of the form 10.
If you press it with a pencil-like device, will it turn into a tablet?
? a is this suppression position [1 is detected and the pressing position is the second designated point 3
Output as position 2. When the operator roughly presses the first specified point 36 corresponding to the upper right corner point of the specified area of the photographic portion 14, and then the second specified point 38 corresponding to the lower left corner point, the tablet 77a detects the pressed position M% and respectively. First and second designated points 36 and 38
Output as the position of . When the operator finishes specifying all specified areas on the form 10, the operator selects the tablet 7? Press the instruction end area 80 on a. When the tablet 77a detects the pressing of the end area 80, it outputs an instruction end signal to the control section 6. As in the first embodiment, when the control section 68 finishes processing all the processing information regarding the input specified point and inputs the instruction end signal, it outputs the positional blindness and processing information of the specified point to the image composition section 66. The image synthesizing section 66 outputs the designated points of the designated areas of the text and photograph portions to the first and second binarization sections 62 and 64. The operator is in the instruction end area 8.
After pressing 0, the form 10 is removed from the input section 77 and set in a scanning mechanism section (not shown). and input section 7
A reading start signal is input to the device of the second embodiment via keypad No. 7. When the reading start signal is input, the photoelectric conversion unit 58 reads the form 1.
A multivalued image pattern of 0 is output and stored in the image memory 60. After inputting the multivalued image pattern of the entire surface of the form 10 into the image memory 60, the first and second binarization units 62 and 64 convert the designated area specified by the position of the specified point input from the image synthesis unit 66. A multivalued image pattern in the image memory 60 is inputted, converted into a simple binary image pattern and a dithered image pattern, and these converted image patterns are outputted to the image synthesis section 66. The image synthesis section 66 synthesizes one image pattern and stores the encoded image pattern in the storage section 70 in the same manner as in the first embodiment. According to this embodiment, similar to the first embodiment, it is possible to improve the operator's work efficiency and provide a device that is easier to operate. Furthermore, since there is no need to provide areas for instructing processing information (document processing instruction area and photo processing instruction area) on the tablet of the input unit, the tablet can be made smaller and the production cost of the device can be reduced. First Rainbow Example This example is an example in which this invention is applied to a character recognition system. FIG. 6 is a functional block diagram schematically showing the configuration of a third embodiment of the present invention. Components similar to those described above are designated by the same reference numerals, and detailed description thereof will be omitted.

As shown in FIG. 6, the apparatus of this embodiment includes a preprocessing section 82 for obtaining a binary image pattern, an image processing section 84 for performing recognition processing, and an input section 56. In this embodiment, the preprocessing section 82 includes a photoelectric conversion section 86 that outputs a binary image pattern and an image memory 60. A recognition unit 88 roughly cuts out a recognition target pattern through the image processing unit 84 and performs recognition processing on the recognition target pattern.
and a control section 90 that detects processing information. The operation of this embodiment will be explained below. In this example, the specified area is a rectangular area connecting the first and second specified points with the line segments as diagonals, and the direction of the vector with the first and second specified points as the starting and ending points is to the right. A rule is predetermined that a horizontally written document recognition process is performed on the specified area specified by the first and second specified points, and when the direction of the vector is leftward, a vertically written document recognition process is performed on the specified area. The operator operates the pretense in this example according to these rules. An operator sets a form (see FIG. 13(A) or (B)) of a horizontally or vertically written document in a scanning mechanism (not shown). The operator then inputs a reading start signal via the mouse. The photoelectric conversion unit 86 starts optical scanning of the form, converts the optical information L into a binary electrical signal, and outputs an image pattern made of this electrical signal. The image memory 60 stores the binary image pattern from the photoelectric conversion unit 86. The control section 90 reads out the binary image pattern from the image memory 60 and transfers it to the display section 72. The display unit 72 displays the image pattern of the entire document based on the transferred image pattern. FIG. 7 is a diagram showing an example of a display screen in this embodiment, and FIGS. 7(A) and (B) show a cursor on the display screen 76. 22 and form patterns 48 and 50 are shown.

When the operator finds a character portion on the display screen 76, the operator determines whether it is a horizontally written document or a vertically written document. If the document is written horizontally, the display position of the cursor 22 is moved to a position corresponding to the upper left corner point of the specified area 28 by operating the mouse. If the document is written vertically, move the cursor 22 to the position corresponding to the upper right corner of the designated area tfi28. When the arrow tip of the cursor 22 reaches a position indicating the desired first designated point 30, the operator operates the mouse button to output a first designated point designation signal. The operating section 74 into which the designated point instruction signal is input detects the arrow tip position of the cursor 22 at this time as the position of the first designated point 30, and outputs it to the control section 90.

Roughly, the operator moves the position of the cursor 22 by scanning the mouse to point to the second specified point 32 in the specified area 28. If the document is written horizontally, the display position of the cursor 22 is moved to the position corresponding to the lower right corner of the specified area 28. If the document is vertically written, the cursor 22 is moved to a position corresponding to the lower left corner point of the specified area 28.

When the arrow tip of the force sol 22 reaches a position that indicates the desired second designated point 32, the operator operates the mouse button to output a second designated point designation signal. The operating unit 74 inputting the specified point instruction signal detects the arrow tip position I8 of the cursor 22 at this time as the position of the second specified point 32, and outputs it to the control unit 90. The control section 90 sequentially inputs the position of the specified point from the operation section 74 and stores it. Position a of the predetermined number of designated points that designate the designated area
When t is input, processing information detection starts. The control unit 90 can detect processing information in parallel with the input of specified points from the input unit 56. The control unit 90 detects the direction of the vector having the specified points 30 and 32 as starting and ending points in the same way as the control unit 68 of the first embodiment, and if it is right-oriented, it detects processing information indicating that horizontally written document recognition processing is to be performed. and if the document is oriented to the left, select processing information indicating that vertical document recognition processing is to be performed, save the 100th place of specified points 30 and 32 along with the selected processing information, and select the specified points 30 and 32. Finish detecting processing information regarding the area. After the operator has designated designated points in all designated areas on the display screen 46, he operates the mouse to output a designated point designation completion signal to the control unit 90.

When the control section 90 finishes processing all the processing information regarding the input designated point and receives an instruction end signal, it outputs the position of the designated point and the processing information to the recognition section 88 . When the recognition unit 88 inputs the specified point position and processing information,
A binary image pattern within the specified area 28 specified by the specified points 30 and 32 is scanned, black pixels on each scanning line are counted to create a first marginal distribution, and a first marginal distribution is created based on the first marginal distribution. to detect the character line position. Furthermore, based on the character line position, the binary image pattern in the character line area is scanned with the main scanning direction perpendicular to the direction along the character line, and the black pixels on the scanning line are counted for each scanning line. A marginal distribution is created, and the position of one unit of recognition target pattern, for example, a character pattern of one character, is detected based on the second marginal distribution.

If the processing information is information representing recognition processing for a horizontally written document, the main scanning direction when creating the first and second peripheral distributions should be the X-axis direction and Y-axis direction, and if the processing information is information representing recognition processing for a vertically written document. Main scanning direction 8 when creating the first and second peripheral distributions
A binary image pattern is scanned in the Y-axis direction and the X-axis direction.

When the recognition unit 88 detects the digit M of one unit of the recognition target, it cuts out the recognition target pattern for each unit from the binary image pattern, extracts various features t that intersect with the cut out recognition target pattern, and diagrams the extracted feature amounts. Recognition is performed by comparing the features of standard characters stored in a dictionary. Then, the recognition result C of the recognition target pattern, for example, the character code specified in the JIS standard, is output to the control group 90. When the control unit 90 receives all the recognition results C within the specified area, it outputs the recognition results C. FIG. 8 is a diagram showing another display example in this embodiment, showing a form pattern 9 in which horizontally written and vertically written character groups are mixed.
2 and cursor 22 are displayed. In the above example, we have explained a form in which all the character groups on the form are in the same character arrangement, but in this embodiment, not only the case where the character groups on the form are the same character arrangement, but also horizontal and vertical character groups are mixed on the same form. It can also be applied to cases. This kind of mixture of horizontal and vertical writing can be seen, for example, in address writing on mail. This section explains the behavior when horizontal and vertical writing are mixed. In this case, the operator inputs to the control unit 90 the specified point instruction signals of the first and second specified points 30 and 32 that specify the specified area 28 of the horizontally written character group, and the specified area of the vertically written character group. The first and second designated points 3 specifying 34
After inputting the specified point instruction signals 6 and 38, input the instruction end signal. Then, the control unit 90 outputs the specified points 30 and 32 of the specified area 28 and processing information, and the specified points 36 and 38 of the specified area 34 and the specified points 36 and 38 and outputs the processing information to the recognition unit 88. The recognition unit performs recognition processing of the binary image pattern within each designated area based on the input designated point position and processing information.

According to this embodiment, processing information that was performed in the conventional layout (whether it is a horizontal writing processing area for recognizing horizontally written documents or a vertical writing processing area for performing recognition processing for vertically written documents) Since the work of inputting information can be omitted, the amount of mouse movement and the number of mouse button operations can be significantly reduced.As a result, it is possible to improve the operator's work efficiency and provide a device that is easy to operate.

Furthermore, in the conventional design, areas for instructing processing information (horizontal writing processing instruction area and vertical writing processing instruction area) were displayed on the display screen, but in this embodiment, these instruction areas can be omitted, and therefore the display device can be made smaller. The production cost of the device can be reduced.

No.! υe Rainbow Example FIG. 9 is a functional block diagram schematically showing the configuration of a fourth embodiment of the present invention. Components corresponding to those described above are denoted by the same reference numerals, and detailed explanation thereof will be omitted.

The device of the fourth embodiment has the same configuration as the third embodiment except that the input unit 77 is composed of an input device called a tablet and a keyboard, but the operation is slightly different from that of the third embodiment.
Hereinafter, points that are different from the third embodiment will be mainly explained, and detailed explanations about the same points as the third embodiment will be omitted. FIG. 10 is a diagram showing a state in which a form is placed on the tablet of the input unit of the fourth embodiment, and FIG. 10 (A) and (B)
) is written in the form 40 and 4 in the predetermined position on the tablet 77a.
2 (see Fig. 13 (A) and CB)) is shown in a fixed state. The operation of the fourth embodiment will be explained with reference to FIG. The operator fixes the form at a predetermined position on the tablet 77a and determines whether the form is a horizontally written document or a vertically written document. If the document is written horizontally, use a pencil-like tool to press N, which corresponds to the upper left corner of the specified area. If it is a vertically written document, press M corresponding to the upper right corner point of the specified area.

The tablet 77a detects this pressed position M and outputs the pressed position as the position of the first specified point 30.

Next, in order to designate the second designated point, the operator presses a position corresponding to the lower right corner of the designated area with a pencil-like instrument if the document is written horizontally. If it is a vertically written document, press 100, which corresponds to the lower left corner point of the specified area. Tablet? ?
a detects this depression position Xt and outputs the depression position 100 as the position of the second specified point 32. When the operator has finished specifying all specified areas, he presses the instruction end area 80. Tablet 77a is the end area ti
The control unit 68 detects the pressing of ili 80 and sends an instruction end signal.
Output to . As in the first embodiment, when the control unit 68 completes all the processing information regarding the input specified point and inputs the instruction end signal, the control unit 68 transfers the position and processing information of the specified point to the recognition unit 8.
The output changes depending on the day.

When the operator finishes pressing the instruction end area 80, the operator sets the form in a scanning mechanism section (not shown) and inputs a reading start signal via the keyboard of the input section 77.

When a reading start signal is input, the photoelectric conversion section 58 outputs a binary image pattern of the form and stores it in the image memory 60. After inputting the multivalued image pattern of the entire surface of the form into the image memory 60, the recognition unit 88 performs recognition processing on the binary image pattern within the specified area specified by the input position of the specified point, and outputs the recognition result. Output to the control unit 90. The control unit 90 outputs the recognition result Ct when the recognition unit 88 completes the recognition process for one sheet. According to this embodiment, as in the third embodiment, processing information (whether it is a horizontal writing processing area for performing recognition processing of horizontally written documents or a vertical writing processing area for performing recognition processing for vertically written documents)
Since the work of inputting information can be omitted, it is possible to improve the operator's work efficiency and provide a device M that is easy to operate. Furthermore, since there is no need to provide areas for instructing processing information (horizontal writing processing instruction area and vertical writing processing instruction area) on the tablet, the tablet can be made smaller, and the production cost of the device can therefore be reduced. The present invention is not limited only to the embodiments described above, and therefore, the input/output signals of each component, the flow of input/output signals, the operation, the flow of the operation, the configuration, numerical conditions, shape, arrangement position, The dimensions and numerical conditions can be changed as desired.

For example, there may be four types of vector directions: diagonally upward and downward to the right, diagonally upward and downward to the left, and different processing information may be associated with each of these direction types. For example, the position of the specified point (x+,'/+
) and the position of the second designated point (X2,'/2), then i
) x, diagonally upward to the right when <x2 and "/+>Vz, i
When D X + < X2 and '/I<'/2, diagonally downward to the right; iii) When x, >
When , it can be determined that the direction is diagonally downward to the left. Furthermore, the number of area designation points is not limited to two, but may be three or more. If there are three or more, processing information can be prepared according to the direction of the vector and the number of specified points, and the processing information can be selected according to the direction of the vector and the number of specified points. Furthermore, if there are three or more, it is possible to specify a specified area having a more complicated shape than in the above embodiment. Furthermore, in the above-described embodiments, an example was explained in which the input unit was configured using a mouse or a tablet, but the input unit may be configured using something other than a mouse or a tablet. For example, the input unit may be configured using a light ben. You can also do this. When using a light ben, a display part is provided in the input part, and the position M on the display screen of the display part specified by the light ben can be input as the place of the area specified point. (Effects of the Invention) As is clear from the above description, according to the image processing device of the present invention, the image processing unit detects processing information of a specified area, so the operator performs the work of specifying a specified point. There is no need to input processing information for the specified area. This improves the work efficiency of the operator and provides an easy-to-operate tool.

[Brief explanation of drawings]

FIG. 1 is a functional block diagram schematically showing the configuration of a first embodiment of the present invention, FIG. 2 is an example of a display screen in the first embodiment, FIG. 3 is an explanatory diagram of processing information detection, and FIG. 4 is a functional block diagram schematically showing the composition of the second embodiment of the present invention; FIG.
FIG. 6 is a functional block diagram schematically showing the configuration of a third embodiment of the present invention, and FIGS. 7(A) to (B) show examples of display screens in the third embodiment. 8 is another display screen example in the third embodiment, FIG. 9 is a functional block diagram schematically showing the configuration of the fourth embodiment of the present invention, and FIG. 10 (A) to (゜B) 11 is a diagram showing a state where a form is loaded on the tablet of the input unit of the fourth embodiment, FIG. 11 is an example of a form containing a mixture of text and photos, and FIG. 12 is an example of a display screen in a conventional document file device. 13(A) to 13(B) are examples of forms for horizontally written documents and vertically written documents, and FIGS. 14(A) to 14(B) are examples of display screens in a conventional character recognition device. 52, 82... Pre-processing section, 54, 84... Image processing section 56, 77... Input section, 58, 86... Photoelectric conversion section 60... Image memory, 62... First Binarization unit 64...Second binarization unit, 66...Image synthesis unit 68, 90...Control unit, 70...Storage unit 7
2...Display section, 74...Operation section 88...
Recognition part. Patent Applicant: Oki Electric Industry Co., Ltd. 18 Form pattern 22 Cursor 24 Character part pattern 26 Photograph part pattern 28, 34 - Designated area 30, 32, 36, 38 Designated point 76 Display screen Second example of display screen in Ichitora's embodiment Figure 3 Functional block diagram of the third example 10. Form 12: Text portion 14 Photo portion 77a Tablet 8.1 Indication end area State of placing the form on the tablet @ Figure 5 Figure 3 An example of a display screen in the English embodiment Figure 7 Example of another display screen in the Example Figure 8 Diagram showing the appearance of a form loaded on a tablet Figure lO Figure 4 Functional block diagram of the example @ 9 Figure Example of a form with a mixture of text and photos Figure 11 Example of a display screen in conventional file mode Figure 12 Example of a form Figure 13

Claims (3)

[Claims] (1) A preprocessing unit for obtaining an image pattern of an information medium;
an image processing unit that performs image processing within a specified area of the image pattern according to processing information, the image processing device comprising: an input unit for detecting the position of a specified point in the specified area on the image pattern; . An image processing device, wherein the image processing unit is an image processing unit that detects the processing information based on at least the position of the specified point. (2) The preprocessing unit is a preprocessing unit that obtains a multivalued image pattern, and the image processing unit is a first binarization unit that performs simple binarization processing of the image pattern; a second binarization unit that performs a binarization process using dither; and an image synthesis unit that synthesizes an image pattern that has been subjected to the simple binarization process and an image pattern that has undergone the binarization process that has been performed using the dither. 2. The image processing apparatus according to claim 1, further comprising: a control unit that detects processing information and encodes a synthesized image pattern; and a storage unit that stores the encoded image pattern. (3) The preprocessing unit is a preprocessing unit that obtains a binary image pattern, and the image processing unit is a recognition unit that cuts out a recognition target pattern from the image pattern and performs recognition processing on the recognition target pattern, and a recognition unit that performs recognition processing on the recognition target pattern, and a processing information 3. The image processing apparatus according to claim 2, further comprising a control section that performs detection.