JP3792402B2 - Image processing apparatus, binarization method, and machine-readable recording medium recording a program for causing a computer to execute the binarization method - Google Patents

Description

[0001]
BACKGROUND OF THE INVENTION
The present invention is used for preprocessing (binarization of images) of scanners, digital copiers, OCRs, etc., and in particular, even a black and white document is processed as a color image, and the back side is halftone on the surface of the document. BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus and a binarization method for obtaining a desired binary image by utilizing the difference between components and color components, and a machine-readable recording medium on which a program for causing a computer to execute the binarization method is recorded .
[0002]
[Prior art]
Conventionally, when a black and white document is input as an image by a scanner or the like, it is first read as a brightness signal and binarized. This binarization technique is disclosed in, for example, Japanese Patent Laid-Open Nos. 6-195511 and 9-305754. Japanese Patent Laid-Open No. 6-195511 discloses a “binarization threshold value determination apparatus” that determines a binarization threshold value by using a density histogram of a multilevel image as an input to a neural network. In “Image processing method” of Japanese Patent Laid-Open No. 9-305754, binarization processing is performed in units of blocks.
[0003]
Further, in consideration of the back-shift in a color image, as a method for determining a black character and a halftone area, Japanese Patent Laid-Open No. 7-226842 “Background control method and method in a color image forming apparatus” is disclosed. This method obtains a halftone such as a color background instead of binarization.
[0004]
[Problems to be solved by the invention]
However, in the techniques disclosed in the conventional Japanese Patent Laid-Open Nos. 6-195511 and 9-305754 as described above, binarization is processed only depending on the brightness value, When the back shining portion appears on the image signal as intermediate brightness, it was difficult to correctly remove this and binarize it.
[0005]
In other words, if the conventional binarization process is performed on a white portion on the front side of the document and a black portion on the back side, “black” is determined depending on the threshold level setting. There was a problem of becoming.
[0006]
The present invention has been made in view of the above. By processing even a black and white original as a color image, a difference occurs between the halftone component and the color component on the back surface of the original. The purpose is to obtain a desired binary image by using it, and to avoid the occurrence of backshifting.
[0007]
[Means for Solving the Problems]
In order to achieve the above object, in the image processing apparatus according to claim 1, image input means for inputting a monochrome image as an RGB color image, and RGB of the portion having the highest brightness in the brightness of the color image. Data detection means for detecting the ratio and the RGB ratio of the portion with the lowest brightness; a ternary means for dividing the brightness into three values of “black”, “white”, and “intermediate”; The RGB ratio is checked. When the RGB ratio of the “intermediate” pixel is close to the RGB ratio of the portion with the lowest lightness, “black”, and when the RGB ratio of the “lightest” portion is close to the RGB ratio , “white” is 2 an intermediate pixel binarizing means value determines, in which Ru comprising a.
[0008]
That is, a black-and-white image is input as a color RGB image, the RGB ratio of the highest brightness portion is recorded as Rw, the RGB ratio of the lowest brightness portion is recorded as Rb, and “black”, “white”, “intermediate” is further recorded as brightness. Is divided into three values, and the RGB ratio of the “middle” pixel is examined. If it is close to Rb, it is “black”, and if it is close to Rw, it is “white”. Can be avoided.
[0009]
In the binarization method according to claim 2, the first step of inputting a black and white image as a color RGB image, finding the portion with the highest brightness, and recording the RGB ratio as Rw, A second step of finding a low lightness portion and recording the RGB ratio as Rb, a third step of dividing the lightness into three values of “black”, “white”, and “middle”, and “middle” pixels A fourth step of obtaining a black-and-white binary image by setting “black” when the RGB ratio of the “intermediate” pixel is close to Rb and “white” when close to Rw; Is included.
[0010]
That is, a black-and-white image is input as a color RGB image, the RGB ratio of the highest brightness portion is recorded as Rw, the RGB ratio of the lowest brightness portion is recorded as Rb, and “black”, “white”, “intermediate” is further recorded as brightness. Is divided into three values, and the RGB ratio of the “middle” pixel is examined. If it is close to Rb, it is “black”, and if it is close to Rw, it is “white”. Can be avoided.
[0015]
According to a third aspect of the present invention, there is provided a machine-readable recording medium in which a program for causing a computer to execute the binarization method according to the second aspect is recorded .
[0016]
That is, by recording a program for causing a computer to execute the binarization method according to claim 2 , it is possible to perform binarization processing by the computer.
[0017]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, an image processing apparatus, a binarization method, and a machine-readable recording medium on which a program for causing a computer to execute the binarization method is recorded will be described in detail with reference to the accompanying drawings.
[0018]
(System configuration)
FIG. 1 is a block diagram illustrating a configuration of an image processing apparatus according to an embodiment. In the figure, reference numeral 101 denotes a CCD-equipped scanner or the like, an image input unit as an image input means for optically reading and inputting a black and white original as a color image (R, G, B signals), and 102 a brightness of the color image. It is a lightness detection unit / RbRw detection unit as a lightness / RGB ratio detection means for obtaining the RGB ratio between the minimum part and the maximum part of the lightness.
[0019]
Reference numeral 103 denotes a ternarization unit as a ternarization unit for ternarizing pixels based on lightness into “black”, “white”, and “intermediate”, and 104 converts the intermediate pixel into binary values of black and white based on the RGB ratio. An intermediate pixel binarization unit serving as an intermediate pixel binarization unit for determination, 105 is a control unit that comprehensively controls the apparatus based on a predetermined control program, and 106 is an input color image, connected component data, and region A data storage unit 107 for storing data and the like is a data communication unit 107 that performs data transmission / reception with an external device.
[0020]
(System operation)
Next, the operation of the image processing apparatus configured as described above will be described. First, a document is set on a scanner (image input unit 101), and the document is exposed and scanned to obtain color images of R, G, and B signals. In addition, an image may be input from the outside via the data communication unit 109.
[0021]
Here, an image reading operation and an example of RGB image capturing will be described with reference to FIGS. FIG. 4 is an explanatory diagram showing an example of reading a black and white original with a light source / CCD. In the figure, 401 is a black and white document, 402 is a light source / CCD, and 403 is a pressure plate. FIG. 5 is an explanatory diagram showing an example in which a black and white document is captured as a multi-value signal by a scanner in the document reading of FIG.
[0022]
When the reading light is not totally reflected on the surface of the original 401 and part of it is transmitted, the transmitted light is reflected by the pressure plate 403 on the original and is obtained as an image signal in the same manner as the reflected light on the original surface.
[0023]
Here, each reflected light and each transmitted light (light transmitted and returned again) are expressed as follows.
Lr: Reflected light on the surface of the original Lrw: Reflected light on the white part Lrg: Reflected light in the middle tone Lrb: Reflected light on the black part Lt: Light transmitted and returned again Ltw: Transmitted light Ltb on the back side of the original with the white part : Transmitted light with black part on the back side of the document.
At this time, the image signal becomes a graph as shown in FIG. In this case, depending on the setting level of the threshold value Th, a determination of “black” is made for a portion where the front side of the document is white and the back side is black.
[0025]
Here, for example, as shown in FIG. 5, when each of RGB takes a value in the range of white 255 to black 0,
White parts on both sides: (200, 202, 192)
Black part of the table: (85, 75, 57)
Table is halftone: (156, 160, 140)
The part of the manuscript table that is white and the back is black: (158, 161, 158)
It is assumed that
[0026]
In this example, the reflected light Lrb and Lrg for the ink on the surface has a lower B value than R and G, respectively. However, although the value of B is low when the both sides of the document are white, the difference is small, and B in the reverse side is not much different from R and G.
[0027]
This is because the reflected light Lrw basically reflects the color of the original paper (strictly, it is slightly deviated from white), so that the influence is exerted on the double-sided white part and the surface white part where the ratio of Lrw is large. And a difference between Lt and Lr due to a change in the color component when passing through the original.
[0028]
Here, in order to set the halftone (Lrg) portion to “black” and the back cover (Lrw + Ltb) portion to “white”, the operation processing of the flowchart shown in FIG. 2 may be performed.
[0029]
In FIG. 2, first, an RGB image is input (S201), a portion having the highest brightness is found, and the RGB ratio is recorded as Rw (S202). Further, a portion with the lowest brightness is found, and the RGB ratio is recorded as Rb (S203). Next, the brightness is simply divided into three values of “black”, “white”, and “intermediate” (S204). For this, a predetermined threshold value may be used, or a method of using a histogram of the whole or local brightness may be used. Subsequently, the RGB ratio of the “intermediate” pixel is examined, and as a result, “black” is selected when close to Rb, and “white” when close to Rw. This is executed for each “intermediate” pixel (S205), and a black and white binary image is obtained (S206).
[0030]
Therefore, by executing the above processing, it is possible to obtain a desired binarization result even for a portion that is difficult to judge only by lightness.
[0031]
By the way, when calculating Rw and Rb, Rw and Rb may be the RGB ratio of the pixel specified based on another index instead of the image with the highest brightness and the low pixel. This operation will be described with reference to the flowchart shown in FIG.
[0032]
In FIG. 3, first, an RGB image is input (S301), and ternarization by brightness (“black”, “white”, “intermediate”) is performed (S302). Further, a pixel representing the white portion is specified, and the RGB ratio is recorded as Rw (S303). Subsequently, a pixel representing the black portion is specified, and the RGB ratio is recorded as Rb (S304). Next, the intermediate value image is binarized by the RGB ratio (S305), and a black and white binary image is obtained (S306).
[0033]
In this way, for the RGB image, ternarization of “black”, “white”, and “intermediate” is performed by the brightness value, and the RGB value of the pixel representing black and the RGB value of the pixel representing white are obtained, By comparing the RGB value with the RGB value of the “intermediate” value pixel, and determining the “backward” portion determined as “intermediate” as “white”, the desired binarization can be achieved even if there is backside change. An image can be obtained.
[0034]
Further, as another method described above, for example, the RGB ratio may be obtained not from the RGB ratio of a specific pixel, but from all or some of the pixels in the black portion or the white portion. In this case, the average of each pixel is taken or the median is taken.
[0035]
The present invention can be realized by software in addition to the embodiment described above. For example, the binarization method described with reference to FIGS. 2 and 3 is programmed, stored in a machine-readable recording medium, and executed on a computer.
[0036]
【The invention's effect】
As described above, according to the image processing apparatus (claim 1) of the present invention, a monochrome image is input as a color RGB image, and the RGB ratio of the highest brightness portion and the RGB ratio of the lowest brightness portion are detected. When the brightness is further divided into three values of “black”, “white”, and “intermediate”, the RGB ratio of the “intermediate” pixel is examined, and the RGB ratio of the “intermediate” pixel is close to the RGB ratio of the lowest brightness portion Is “black”, “white” when close to the RGB ratio of the maximum brightness portion, and by obtaining a black and white binary image, an apparatus capable of avoiding the occurrence of a back-slip image is realized.
[0037]
According to the binarization method of the present invention (claim 2), a black and white image is input as a color RGB image, the RGB ratio of the highest brightness portion is recorded as Rw, and the RGB ratio of the lowest brightness portion is set to Rb. Is further divided into three values, “black”, “white”, and “intermediate”, and the RGB ratio of the “intermediate” pixel is examined. If it is close to Rb, it is “black”, and if it is close to Rw, “ As a “white”, a black and white binary image is obtained, and it is possible to avoid the occurrence of a back-slip image.
[0040]
According to the machine-readable recording medium of the present invention (Claim 3 ), a program for causing a computer to execute the binarization method according to Claim 2 is recorded, whereby binarization processing is performed by the computer. Can be done.
[Brief description of the drawings]
FIG. 1 is a block diagram showing a configuration of an image processing apparatus according to an embodiment of the present invention.
FIG. 2 is a flowchart showing a first processing operation example until a monochrome binary image is obtained from an RGB image according to the embodiment.
FIG. 3 is a flowchart showing a second processing operation example until a monochrome binary image is obtained from the RGB image according to the embodiment.
FIG. 4 is an explanatory diagram showing an example of reading a black and white original by a light source / CCD.
5 is an explanatory diagram illustrating an example in which a black and white document is captured as a multi-value signal by a scanner in the document reading of FIG.
[Explanation of symbols]
101 Image Input Unit 102 Brightness Detection Unit / RbRw Detection Unit 103 Trinization Unit 104 Intermediate Pixel Binarization Unit

Claims (3)

Image input means for inputting a black and white image as an RGB color image;
Data detection means for detecting the RGB ratio of the portion with the highest brightness in the brightness of the color image and the RGB ratio of the portion with the lowest brightness;
The ternary means for dividing into three values of “black”, “white”, and “intermediate” according to the lightness, and the RGB ratio of the “intermediate” pixel are examined, and the RGB ratio of the “intermediate” pixel has the lowest lightness. If close to the portion of the RGB ratio "black", is close to the RGB ratio of the portion having a high the most brightness and intermediate pixel binarizing means for determining two values "white",
The image processing device characterized by Ru with a. Inputting a black and white image as a color RGB image, finding a portion with the highest brightness, and recording the RGB ratio as Rw;
A second step of finding the lightest part and recording the RGB ratio as Rb;
A third step of dividing the brightness into three values of “black”, “white”, and “intermediate”;
The RGB ratio of the “intermediate” pixel is examined. When the RGB ratio of the “intermediate” pixel is close to the Rb, it is “black”, and when the RGB ratio is close to the Rw, the white ratio is obtained. 4 processes,
A binarization method comprising: A machine-readable recording medium on which a program for causing a computer to execute the binarization method according to claim 2 is recorded.

Images