Image analysis – Image segmentation – Distinguishing text from other regions
Reexamination Certificate
1999-05-27
2002-10-15
Au, Amelia M. (Department: 2623)
Image analysis
Image segmentation
Distinguishing text from other regions
C382S270000, C358S466000
Reexamination Certificate
active
06466693
ABSTRACT:
BACKGROUND OF THE INVENTION
(1). Field of the Invention
The present invention relates to an image processing apparatus which serves for digital copying machines and facsimile machines and the like, and performs optimum processing according to the properties of an area obtained by scanning an original document for improving the image quality of recorded images.
(2). Description of the Related Art
With an image processing apparatus such as a digital copying machine and a facsimile apparatus and the like, the image quality has been conventionally improved by discriminating screened half tone, continuous tone and a character area included in an original document, respectively, subjecting the screened halftone area to a smoothing processing to suppress a moiré, and subjecting the character area to an emphasizing processing to display characters clearly. To improve the image quality, various image discrimination methods have been used.
As a representative image discrimination processing method, there is known a method for dividing an image into respective predetermined blocks and discriminating each area by using a characteristic parameter representing properties of character and screened halftone images.
As the related art for discriminating character and screened halftone areas in each predetermined block by using a characteristic parameter, for example, some are disclosed in Japanese Unexamined Patent Publication No. Hei 5-167842 and Japanese Unexamined Patent Publication No. Hei 5-41796.
With the “image processing apparatus” disclosed in Japanese Unexamined Patent Publication No. Hei 5-167842, the image quality is improved by comparing the maximum density difference in a predetermined area including a target pixel with a fixed threshold to judge the area, and changing over the post-stage processing. It utilizes a fact that when a black letter is seen in a local area, the maximum density difference which is a difference between the maximum density and the minimum density is larger than the continuous tone area.
Moreover, the “copying apparatus” disclosed in Japanese Unexamined Patent Publication No. Hei 5-41796 improves the image quality by extracting a non-background color pixel using an average value in an N×N window including a target pixel as a threshold, and judging whether the target pixel is a character area or not by the pattern of the non-background color pixel to change over the post-stage processing. It utilizes a fact that, in the character area, the length of the same value of the data after binary coding becomes long, and designates the maximum density difference, a run length which is the length of the same value of the data and the like as the characteristic information to compare the characteristic information and the fixed threshold, and when the characteristic information is larger sag than the fixed threshold, it is discriminated to be a character area, and in other cases, it is discriminated to be a photograph area and the like.
When a black letter in a predetermined block is a small letter, the density difference may not be detected clearly depending upon the precision of a scanner, and there may be a case where even if it is a character area, the maximum density difference becomes small. Therefore, with the threshold processing method using a fixed threshold, erroneous separation may be caused, which becomes a significant problem.
Furthermore, when an area in which a character is included is binarized and used as a run length to designate the maximum continuous value of the binary coding result as a characteristic value, in general, binary coding is performed by using an average value in a predetermined area. However, due to the average value determined in a local area, even where the density change is small, there is a possibility of being divided into erroneous binary values by the binary coding processing, which has been a problem.
SUMMARY OF THE INVENTION
In view of the foregoing problem, it is an object of the present invention to provide an image processing apparatus which discriminates in which area of the character area, the continuous tone area and the screened halftone area each pixel belongs, with respect to an image signal obtained by two-dimensionally scanning an original document in the main scanning direction and in the sub-scanning direction, and performs a processing for improving the image quality to enhance the detection accuracy of, in particular, black letters.
With a view to attaining the above objects, the gist of the present invention is as follows.
The first gist of the present invention is an image processing apparatus which discriminates an area having different properties with respect to an image signal obtained by scanning an original document, and performs adequate processing to each area, wherein said image processing apparatus comprises:
dividing means for dividing the image signal into local blocks having a predetermined size; and
local characteristic value-computing means having:
a binary coding processing section which converts an image signal in a predetermined block including a target pixel into binary code information, from image signals output from the dividing means in an area to be processed;
a threshold processing section which processes characteristic information in the predetermined block with a threshold determined by selectively using each area (dimension) value of binary values from the binary code information output from the binary coding processing section and number of times that the binary code information is changed; and
a judgement processing section which judges whether the target pixel in the predetermined block is a character area or not based on the processing result of the threshold processing section.
The second gist of the present invention is an image processing apparatus which discriminates an area having different properties with respect to an image signal obtained by scanning an original document, and performs adequate processing to each area, wherein said image processing apparatus comprises:
dividing means for dividing the image signal into local blocks having a predetermined size; and
local characteristic value-computing means having:
an edge detection section for detecting an edge in a local block including a target pixel in the local block output from the dividing means;
a binary coding processing section which converts an image signal in a predetermined block including a target pixel into binary code information, from image signals output from the dividing means in an area to be processed; a threshold processing section which processes characteristic information in the predetermined block with a threshold determined by selectively using each area (dimension) value of binary values from the binary code information output from the binary coding processing section and number of times that the binary code information changed; and a judgement processing section which judges whether the target pixel is a character area or not based on the processing result of the threshold processing section; and
a color judgement processing section for judging whether the target pixel which is an edge is a monochrome edge or a color edge, based on the edge detection results of the edge detection section.
The third gist of the present invention is an image processing apparatus which discriminates an area having different properties with respect to an image signal obtained by scanning an original document, and performs adequate processing to each area, wherein said image processing apparatus comprises:
dividing means for dividing the image signal into local blocks having a predetermined size; and
local characteristic value-computing means having:
an edge detection section for detecting an edge in a predetermined block including a target pixel in the predetermined block output from the dividing means;
a binary coding processing section which converts an image signal in a predetermined block including a target pixel into binary code information, from image signals output from the dividing means in an area t
Adachi Yasushi
Kanata Toshihiro
Otsu Makoto
Au Amelia M.
Bhatnagar Anand
Conlin David G.
Dike, Bronstein, Roberts & Cushman IP Group Edwards & Angell
Sharp Kabushiki Kaisha
LandOfFree
Image processing apparatus does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Image processing apparatus, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Image processing apparatus will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2971799