JPH06215179A - Method and device for processing image - Google Patents

Abstract

PURPOSE:To display the estimated result of a composition direction before recognition processing is started by inputting a direction instruction signal and performing control so as to recognize inputted image information according to an instruction based on a composition change instruction signal. CONSTITUTION:Images are read from a scanner 101, and the image information is converted to binary electric signals by a scanner interface circuit 102. The read data are indicated on a display 108. Next, an operator sets a recognition object area by using a pointing device 103. The composition direction is estimated to that area, information whether the composition direction is vertical or horizontal is provided, and that information is held in a RAM 107. The operator judges the estimation of the composition direction due to the device by observing a composition direction mark 52 displayed according to composition direction information. When the indication of the composition direction is proper, the operator clicks a recognition button with the pointing device 103. When it is judged, character segmentation processing is performed to segment a character image for each character.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a character image processing method and apparatus for reading a document image and converting it into a character code.

[0002]

2. Description of the Related Art The process flow of a conventional optical character recognition device will be described with reference to the flowchart of FIG. In step S701, an image is read from the scanner and the image information is converted into a binary electric signal. The read data is displayed on the display in S702. Next, in step S703, the operator sets the recognition target area as a rectangle having an arbitrary size and position using a pointing device. Since there are vertical writing and horizontal writing in Japanese documents, it is necessary to determine the attribute of vertical writing or horizontal writing, that is, the writing direction before the character segmentation processing. Therefore, S704
Then, the set direction is estimated with respect to the area, and in accordance with the result, a character cutting process for extracting a character image for each character is performed in S705, and a recognition process is performed in S706. The recognition result is displayed on the display in step S707, and edited or saved according to the operator's instruction.

[0003]

It is very difficult to estimate the set direction depending on the input document, and an incorrect estimation may be performed in S704. However, in the above-mentioned conventional example, since there is no means for the operator to confirm and correct when the set direction is erroneously estimated, the processing is continued with the wrong set direction and an incorrect recognition result is generated. There was a flaw.

[0004]

In order to solve the above problems, the present invention is an image processing for inputting image information, determining the set direction of the input image information, and notifying the determined set direction. Methods and apparatus are provided.

In order to solve the above problems, the present invention preferably inputs a direction change instruction signal, and controls so as to recognize the input image information in accordance with an instruction by the pair change instruction signal.

[0006]

EXAMPLE A first example of the present invention will be described in detail. FIG. 1 is a block diagram showing the configuration of the first embodiment. In FIG. 1, 101 is a scanner for irradiating an image original with light and reading the reflected light to convert it into an electric signal.
Reference numeral 2 denotes a scanner interface circuit for converting the electric signal obtained at 101 into a binary digital electric signal and transmitting the binary digital electric signal to other device components. Reference numeral 103 denotes a pointing device for inputting coordinates on the window of the display.
Reference numeral 104 is a pointing device interface circuit for transmitting a signal from the pointing device 103 to other device components, and 105 is a CPU for controlling the entire device and executing a character cutting process and a recognition process according to a program stored in a ROM. Reference numeral 106 denotes a ROM that stores a control program for processing executed by the CPU 105, various processing programs, a recognition dictionary, and the like, as shown in a flowchart to be described later. Reference numeral 107 denotes a work area for developing character images and character recognition processing. RAM used, 108 is a display for displaying an input image and a recognition result, 109 is a display interface circuit, and 110 is a CP for connecting each device constituent element.
It is a U-bus.

FIG. 2 is a flow chart showing the flow of processing of this embodiment. In step S201, an image is read from the scanner 101, and the image information is converted into a binary electric signal by the scanner interface circuit 102. The read data is displayed on the display 108 in S202. Next, in S203, the operator sets the recognition target area using the pointing device 103. Since there are vertical writing and horizontal writing in Japanese documents, it is necessary to determine the writing direction before performing the recognition process. Therefore, in S204, the set direction is estimated for the area by a method described later, and set direction information indicating whether the set direction is vertical writing mode or horizontal writing mode is obtained.
It is retained in M107. In S206, the set direction mark 52 is displayed in the lower left corner of the target area on the display 108 as in the display example of FIG. 5 according to the set direction information. The operator looks at this assembly direction mark 52 to judge the estimation of the assembly direction by the device. If the display of the set direction is correct, the operator clicks the recognition button with the pointing device 103. When it is determined in S207, the process proceeds to S210,
In S210, a character cutting process for extracting a character image for each character is performed, and a recognition process is performed in S211. The recognition result is displayed on the display in S212, and is edited or saved according to the operator's instruction.

However, it is very difficult to estimate the set direction depending on the input document, and it is not always possible to accurately estimate the set direction. Therefore, when the estimation of the set direction is incorrect, the operator clicks the set direction mark on the display with the pointing device 103.
As a result, the set direction information is inverted in S209. Upside-down means that until now horizontal writing is vertical writing, and vertical writing is horizontal writing. Then, the set direction mark is re-displayed in S209 according to the reversed set direction information. After that, the process returns to S205, and the subsequent processing is the same.

The method of estimating the set direction performed in S204 in this embodiment will be described with reference to FIG. The x-axis is taken in the horizontal direction with respect to the image data, and the projection component of the black pixel on the x-axis is P (x). Similarly, the y-axis is taken in the vertical direction, and the projection component on the y-axis is P (y). The shaded portion in FIG. 3 represents the magnitude of the projected component.

Next, the variances of P (x) and P (y) are calculated by the following equations.

[0011]

[Outer 1] Is the average of P (x) in this section, and sums all projection components in this section. Further, n is the number of data in this section.

Similarly, the variance of P (y) is also calculated.

[0013]

[Outside 2] The set direction is estimated using Sx and Sy. That is, since the variance of the projection component is large in the direction along the set direction, if Sx> Sy, it is estimated that horizontal writing is performed, and otherwise, vertical writing is estimated.

Further, display examples of the image data and the set direction mark are shown in FIGS. FIG. 4 is an example when the image data is displayed on the display 108. Reference numeral 41 is a window for displaying an image, and “vecog” 42 is a recognition button for instructing recognition execution. FIG. 5 is a diagram showing that the recognition target area 51 is set by the pointing device 103. As soon as the area setting is completed, the set direction is estimated, and the result is the set direction mark 5.
Indicated by 2. This mark represents horizontal writing mode. Reference numeral 61 in FIG. 6 is a set direction mark indicating vertical setting.
However, in this case, the estimation of the set direction is incorrect.

The present invention is effective without any particular limitation on the method of estimating the set direction. Also, it does not depend on the design, display position, color, etc. of the assembly direction mark. Further, the means for reversing the set direction information may be a method such as double-clicking with a mouse or grabbing (dragging) a mark for rotation.

[0016]

As described above, according to the present invention, the set direction estimation result can be displayed before the recognition processing is started, and the set direction information can be easily inverted if it is incorrect. Since it is provided with the means, there is no risk that the processing will be performed erroneously until the end, the overall operation time will be shortened, and a better user interface can be provided.

[Brief description of drawings]

FIG. 1 is a block diagram showing a configuration of an embodiment.

FIG. 2 is a flowchart showing the flow of processing of the embodiment.

FIG. 3 is a diagram for explaining a projected component of a character image.

[Fig. 4] Display example of image data on display

FIG. 5: Display example of set direction mark

FIG. 6 is a display example of a set direction mark.

FIG. 7 is a flowchart of a conventional example.

Claims (4)

[Claims] 1. An input means for inputting image information, a set direction determining means for determining a set direction of the input image information, and a notifying means for notifying the determined set direction. Image processing device. 2. A control means having a set direction change instruction means, and a control means for controlling so as to recognize the image information input by the input means in accordance with an instruction from the set direction change instruction means. The image processing device according to item 1. 3. An image processing method comprising inputting image information, determining a set direction of the input image information, and notifying the determined set direction. 4. The image processing method according to claim 3, wherein a set direction changing instruction signal is input, and control is performed such that the input image information is recognized in accordance with an instruction by the set changing instruction signal.