Information processing system, information processing method, and program

By differentially compressing text and non-text areas and performing preprocessing to enhance image clarity and remove noise, the system addresses the issue of reduced character recognition accuracy in high-compression PDF files, achieving improved recognition accuracy.

JP2024165431AInactive Publication Date: 2024-11-28CANON MARKETING JAPAN INC +1
View PDF 1 Cites 0 Cited by

Patent Information

Application Number
JP2023081634
Authority / Receiving Office
JP · JP
Patent Type
Applications
Current Assignee / Owner
Filing Date
2023-05-17
Publication Date
2024-11-28
Estimated Expiration
Not applicable · inactive patent

AI Technical Summary

Technical Problem

Existing high-compression PDF technologies can mistakenly compress characters as background, leading to reduced accuracy in character recognition processing.

Method used

The system differentially compresses text and non-text areas, extracts character strings from non-text areas, and performs preprocessing to invert pixel colors and remove noise, enabling accurate character recognition without requiring uncompressed data.

Benefits of technology

This approach effectively suppresses the decrease in character recognition accuracy for highly compressed PDF files by enhancing image clarity and reducing noise, ensuring accurate character extraction.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure 2024165431000001_ABST
    Figure 2024165431000001_ABST
Patent Text Reader

Abstract

To prevent a reduction in the accuracy of character recognition on an input image file without requiring image data before compression.SOLUTION: An information processing apparatus comprises: acquisition means that acquires data obtained by compressing a character area and a non-character area included in data in different methods; extraction means that extracts a character string from an area compressed as the non-character area; and processing means that performs predetermined processing on the character string extracted by the extraction means.SELECTED DRAWING: Figure 1
Need to check novelty before this filing date? Find Prior Art

Description

[Technical field]

[0001] The present invention relates to an information processing system, an information processing method, and a program. [Background technology]

[0002] When scanning documents such as forms, high-compression PDF technology exists as a technique for reducing file size while maintaining image quality.

[0003] High-compression PDF technology separates a scanned image into a text image and a background image, and then combines them after applying appropriate compression processing to each, making it possible to compress the file size while maintaining the readability and clarity of the text.

[0004] However, some characters may be mistakenly determined to be background (non-character), resulting in inappropriate compression of the character image, which may result in some characters being noisy and reducing the accuracy of character recognition processing.

[0005] Patent Document 1 discloses a technique for suppressing the degradation of character recognition accuracy caused by highly compressed PDF files. [Prior art documents] [Patent documents]

[0006] [Patent Document 1] JP 2018-018361 A DISCLOSURE OF THEINVENTION [Problem to be solved by the invention]

[0007] Patent Document 1 describes that character recognition processing is performed by distinguishing between character regions and non-character regions by using compressed data and uncompressed data.

[0008] In addition, it is described that if the separation of character and non-character areas is inappropriate, degradation of recognition accuracy can be suppressed by performing character recognition processing on the character area data before compression.

[0009] However, the technique of Patent Document 1 requires uncompressed data in order to prevent a decrease in recognition accuracy.

[0010] SUMMARY OF THE PRESENT EMBODIMENT An object of the present invention is to provide a mechanism capable of suppressing deterioration in accuracy in character recognition of an input image file without requiring uncompressed image data. [Means for solving the problem]

[0011] an acquisition means for acquiring data obtained by compressing character areas and non-character areas included in the data using different compression methods; an extraction means for extracting a character string from the area compressed as the non-character area; a processing means for performing a predetermined process on the character string extracted by the extraction means; An information processing device comprising: Effect of the Invention

[0012] According to the present invention, it is possible to suppress a decrease in accuracy in character recognition of an input image file without requiring image data before compression. [Brief description of the drawings]

[0013] [Figure 1] FIG. 1 is a diagram illustrating an example of a system configuration according to an embodiment of the present invention. [Diagram 2] FIG. 2 is a block diagram showing an example of a hardware configuration of a PC in an embodiment of the present invention. [Diagram 3] 10 is a flowchart showing a series of processes for an input PDF file in the embodiment of the present invention. [Figure 4] 5 is a flowchart showing a process flow of a high-compression PDF determination unit in the embodiment of the present invention. [Diagram 5] 3A to 3C are diagrams illustrating an example of images constituting a highly compressed PDF file according to an embodiment of the present invention. [Figure 6] 5 is a flowchart showing a processing flow of an image pre-processing unit in the embodiment of the present invention. [Figure 7] 11A and 11B are diagrams illustrating an example of a correction result of a mask image in the embodiment of the present invention. [Figure 8] 11A and 11B are diagrams illustrating an example of a correction result of a background image in the embodiment of the present invention. [Figure 9] FIG. 11 is a diagram showing an example of an image synthesis result in the embodiment of the present invention. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0014] FIG. 1 is a diagram showing an example of a system configuration of a handwritten character extraction system according to an embodiment of the present invention.

[0015] A client terminal 101 that accepts operations by a user and a scanner 102 that scans forms and creates image files are connected via a communication path 100 .

[0016] The communication path 100 can take the form of a wired LAN, a wireless LAN, a USB, etc., depending on the physical interface of the scanner 102 .

[0017] The server device 103 has a function of storing image data scanned by the scanner 102 and performing OCR processing on the image data based on an instruction from the client terminal 101 .

[0018] It should be noted that image data storage and OCR processing may be performed by the client terminal 101 instead of the server device 103 .

[0019] FIG. 2 is a block diagram showing an example of a hardware configuration of an information processing device applicable as the client terminal 101 or the server device 102 of the present invention.

[0020] As shown in FIG. 2, the information processing device has a CPU (Central Processing Unit) 201, a ROM (Read Only Memory) 202, a RAM (Random Access Memory) 203, a storage device 204, an input controller 205, an audio controller 206, a video controller 207, a memory controller 208, and a communication I / F controller 209 connected via a system bus 200.

[0021] The CPU 201 comprehensively controls each device and controller connected to the system bus 200 .

[0022] ROM 202 or external memory 213 holds the BIOS (Basic Input / Output System) and OS (Operating System), which are control programs executed by CPU 201, computer-readable and executable programs for realizing this information processing method, and various necessary data (including data tables).

[0023] The RAM 203 functions as a main memory, a work area, etc., for the CPU 201. The CPU 201 loads programs and the like required for executing processes from the ROM 202 or the external memory 213 into the RAM 203, and executes the loaded programs to realize various operations.

[0024] The input controller 205 controls input from an input device such as a keyboard 210 or a pointing device such as a mouse (not shown). If the input device is a touch panel, the user can give various instructions by pressing (touching with a finger or the like) icons, cursors, or buttons displayed on the touch panel.

[0025] The touch panel may be a touch panel capable of detecting positions touched by multiple fingers, such as a multi-touch screen.

[0026] The video controller 207 controls display on an external output device such as a display 212. The display also includes the display of a notebook computer integrated with the main body. Note that the external output device is not limited to a display, and may be, for example, a projector. In addition, for the device capable of receiving the above-mentioned touch operation, an input device is also provided.

[0027] The video controller 207 is capable of controlling a video memory (VRAM) for display control, and can use part of the RAM 203 as the video memory area, or can provide a separate dedicated video memory.

[0028] The memory controller 208 controls access to the external memory 213. As the external memory, an external storage device (hard disk) for storing a boot program, various applications, font data, user files, edit files, and various data, a flexible disk (FD), or a compact flash (registered trademark) memory connected to a PCMCIA card slot via an adapter, etc., can be used.

[0029] The communication I / F controller 209 connects and communicates with external devices via a network, and executes communication control processing on the network. For example, communication using TCP / IP, telephone lines such as ISDN, and communication using 4G and 5G lines for mobile phones are possible.

[0030] The CPU 201 enables display on the display 212 by, for example, executing a process of expanding (rasterizing) an outline font in a display information area in the RAM 203. The CPU 201 also enables a user to give instructions using a mouse cursor (not shown) or the like on the display 212.

[0031] Next, the processing executed by the information processing system according to the embodiment of the present invention will be described with reference to the flowcharts of FIGS. FIG. 3 is a flowchart showing an overall series of processing steps according to the present invention, in which the CPU 201 reads out and executes a predetermined control program.

[0032] In step S301, the PDF file input to the input receiving unit 251 is obtained.

[0033] In step S302, it is determined whether the PDF file acquired in step S301 is a highly compressed PDF file or a non-highly compressed PDF file (hereinafter, a non-highly compressed PDF file).

[0034] In step S303, it is checked whether the PDF file determined in step S302 is a highly compressed PDF file. If it is a highly compressed PDF file, the process proceeds to step S304. If it is a non-highly compressed PDF file, the process proceeds to step S305.

[0035] In step S304, a predetermined pre-processing is performed on the PDF file determined to be a highly compressed PDF file in step S303. In the present invention, the pre-processing includes removing mosquito noise and binarization. By performing appropriate pre-processing on a highly compressed PDF file, it is possible to suppress deterioration in the accuracy of character recognition.

[0036] In step S305, character recognition processing is performed on the highly compressed PDF file that has been preprocessed in step S304 or the non-highly compressed PDF file sent from step S303. In the character recognition processing, a deep learning model such as a Convolutional Recurrent Neural Network (CRNN) may be used.

[0037] In step S306, the result output unit 255 outputs the character recognition result obtained in step S305. FIG. 4 is a flowchart showing the details of step S302.

[0038] In step S401, the PDF file acquired in step S301 is analyzed to acquire image information. Here, image information refers to information obtained by analyzing the pages of the PDF file. Examples of the image information obtained include the size, color space, and coordinate system of a single or multiple images that make up each page.

[0039] In step S402, the number of images constituting the pages of the PDF file acquired in step S301 is checked from the image information acquired in step S401, and it is determined whether the target PDF file is a highly compressed PDF file.

[0040] The images that make up a highly compressed PDF file will be described using the example shown in FIG. 5. In a highly compressed PDF file, a page image (501) is separated into a background image 502 and a mask image 503 and saved. A background image is an image in which only areas determined to be non-text are extracted from the original image. A mask image is an image in which only areas determined to be text are extracted from the original image, and is saved with the background in black and the text in white. On the other hand, in the case of a non-highly compressed PDF file, a page image is saved without being separated. In this way, by checking the number of images that make up a page, it is possible to determine whether the target PDF image file has been highly compressed.

[0041] If the page of the PDF file obtained in step S301 contains multiple images, the process proceeds to step S403, where the target PDF file is determined to be a highly compressed PDF image. If the page contains a single image, the process proceeds to step S404, where the target PDF file is determined to be a non-highly compressed PDF image. FIG. 6 is a flowchart showing details of the process in step S304.

[0042] In step S601, it is checked whether an image included in a page of the PDF file determined to be a highly compressed PDF file in step S302 is a mask image. One method for determining whether an image is a mask image is, for example, to check the ratio of black pixels to white pixels in the entire image. Since a mask image has a black background and white characters, if there are significantly more black pixels than white pixels, the target image can be determined to be a mask image. If it is a mask image, the process proceeds to step S602, where a predetermined process is performed. If it is not a mask image, the target image is determined to be a background image, and a series of processes are performed in steps S603 to S606.

[0043] In step S602, the image (503) determined to be a mask image in step S601 is subjected to a process of inverting the white and black colors of each pixel of the image. This results in a corrected mask image 801 (FIG. 7). By performing the inversion process, the background and character colors become the same as those of a corrected background image 901 (described later), making it possible to create a composite image suitable for character recognition in step S607 (described later).

[0044] In step S603, character detection processing is performed on the background image 502 that is determined not to be a mask image in step S601. As a means for character detection, for example, a deep learning model such as FCENet (Fourier Contour Embedding Network) may be used. By performing the character detection processing, a rectangle of a character or a character string and its coordinate system are obtained. For example, in FIG. 8, character regions 505 to 507 are obtained as a result of character detection.

[0045] In step S604, the character regions are extracted from the background image 502 based on the rectangular coordinates of the character regions 505 to 507 obtained in the character detection process in step S603.

[0046] In step S605, mosquito noise is removed from each character region extracted in step S604. Mosquito noise here refers to small dot-like noise that appears around the outlines of characters in highly compressed images. As a means for removing mosquito noise, an algorithm such as Non-Local Means Denoising may be used.

[0047] In step S606, the result obtained in step S605 is subjected to binarization processing. Here, binarization is processing to convert the target image into a binary image consisting of only black and white. Through the above processing, a corrected background image 901 (FIG. 8) is obtained.

[0048] In step S607, a process is performed to combine the corrected mask image 801 obtained in step S602 and the corrected background image 901 obtained in step S606. The corrected mask image 801 is embedded in the corrected background image 901 based on the coordinate information of each image obtained in step S401. This results in a composite image 1001 (FIG. 9). In the composite image 1001, noise has been removed by preprocessing, so characters are clearer than in a highly compressed PDF file that is not preprocessed. Therefore, in the character recognition process in step S305, it is possible to suppress the decrease in accuracy compared to a highly compressed PDF file that is not preprocessed.

[0049] Although the embodiments have been described above, the present invention can be embodied, for example, as a system, an apparatus, a method, a program, or a recording medium, etc. Specifically, the present invention may be applied to a system composed of multiple devices, or may be applied to an apparatus composed of a single device.

[0050] Moreover, the program of the present invention is a program that enables a computer to execute the processing methods of the flowcharts shown in Figures 3, 4 and 6, and the storage medium of the present invention stores a program that enables a computer to execute the processing methods of Figures 3, 4 and 6. The program of the present invention may be a program for each processing method of each device in Figure 3.

[0051] As described above, it goes without saying that the object of the present invention can be achieved by supplying a recording medium on which a program that realizes the functions of the above-mentioned embodiments is recorded to a system or device, and having the computer (or CPU or MPU) of that system or device read and execute the program stored on the recording medium.

[0052] In this case, the program itself read out from the recording medium will realize the novel functions of the present invention, and the recording medium on which the program is recorded will constitute the present invention.

[0053] Examples of recording media for supplying the program include flexible disks, hard disks, optical disks, magneto-optical disks, CD-ROMs, CD-Rs, DVD-ROMs, magnetic tapes, non-volatile memory cards, ROMs, EEPROMs, silicon disks, etc.

[0054] Furthermore, it goes without saying that the functions of the above-mentioned embodiments are not only realized by the computer executing a program it has read, but also includes cases where an OS (operating system) running on the computer performs some or all of the actual processing based on the instructions of the program, and the functions of the above-mentioned embodiments are realized through that processing.

[0055] Furthermore, it goes without saying that this also includes cases where a program read from a recording medium is written into a memory provided on a function expansion board inserted into a computer or a function expansion unit connected to a computer, and then a CPU or the like provided on the function expansion board or function expansion unit performs some or all of the actual processing based on the instructions of the program code, thereby realizing the functions of the above-mentioned embodiments.

[0056] The present invention may be applied to a system consisting of multiple devices, or to an apparatus consisting of a single device. Needless to say, the present invention can also be applied to a case where the object is achieved by supplying a program to a system or apparatus. In this case, the effect of the present invention can be enjoyed by the system or apparatus by reading a recording medium storing a program for achieving the present invention into the system or apparatus.

[0057] Furthermore, by downloading and reading a program for achieving the present invention from a server, database, etc. on a network using a communication program, the system or device can enjoy the effects of the present invention. Note that the present invention also includes configurations that combine the above-mentioned embodiments and their modified examples. [Explanation of symbols]

[0058] Client terminal 101 Scanner 102 Server device 103

Claims

1. an acquisition means for acquiring data obtained by compressing character areas and non-character areas included in the data using different compression methods; an extraction means for extracting a character string from the area compressed as the non-character area; a processing means for performing a predetermined process on the character string extracted by the extraction means; An information processing device comprising:

2. 2. The information processing apparatus according to claim 1, wherein the processing by said processing means is a process for removing noise.

3. 3. The information processing apparatus according to claim 2, wherein the processing by said processing means is a binarization process.

4. 4. The information processing apparatus according to claim 1, further comprising an OCR unit for performing an OCR process on the data processed by the processing unit.

5. an acquiring step of acquiring data obtained by compressing character regions and non-character regions included in the data using different compression methods; an extraction step of extracting a character string from the region compressed as the non-character region; a processing step of performing a predetermined process on the character string extracted by the extraction means; An information processing method comprising:

6. A program for causing a computer to function as each of the means according to any one of claims 1 to 4.

Citation Information

Patent Citations

  • Data processor and computer program

    JP2018018361A