JP2003108997A - Threshold setting method for binarization process, binarization processing circuit, binarization processing program and computer-readable recording medium - Google Patents

Abstract

PROBLEM TO BE SOLVED: To make automatically settable a threshold for a binarization process on the basis of a desired area, and to make settable the plurality of thresholds. SOLUTION: The binarization process is executed to an input image on the basis of a prescribed threshold. The binarization processing method is provided with a process for displaying data on the input image as an image; a process for generating a density histogram on the basis of the input image, a process for calculating the density threshold value on the basis of a characteristic quantity expressed by peak numbers of peaks included in the density histogram of the input image and an offset amount of the peak number from a peak position, a process for specifying an arbitrary position to the displayed image by a user, and a process for setting the density threshold on the basis of the characteristic quantity obtained from a density value of the position specified on the image by the user.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for determining a reference threshold value, a binarization processing circuit, a binarization processing program, and a computer in the binarization processing performed in the preprocessing of image processing. The present invention relates to a recording medium that can be read.

[0002]

2. Description of the Related Art Image processing is used in various fields such as manufacturing and mounting of electronic parts in factories. In the image processing, a binarization processing for binarizing the image data based on a predetermined threshold value is performed as a preprocessing. In the binarization process, it is important to determine the threshold value.

Conventionally, when performing image processing using a density threshold value, a discriminant analysis method has been used as a method for determining the threshold value.
The modal method was used. In these methods, one density is automatically calculated from the density histogram of an image and the value is used as a threshold. The density histogram shows how many pixels of each density exist in the input image.

[0004]

However, these methods have a problem that the threshold value cannot be set by designating a specific portion because the threshold value is calculated from the entire image. The user may want to specify a region or density value of particular interest in the image and perform binarization with a threshold value based on the specified information. However, in the conventional method, the threshold value cannot be calculated based on the partial area, and only the threshold value can be obtained from the entire image.

Furthermore, the conventional method has a problem that only one threshold value can be determined. For this reason, it is not possible to set a plurality of optimum threshold values for each region designated by the user, and only one threshold value based on the entire image can be calculated.

The present invention was developed to solve such problems. An object of the present invention is to automatically calculate a threshold value based on a specific part or a desired region, and further set a plurality of threshold values in a threshold value setting method in binarization processing, and binarization. It is to provide a processing circuit, a binarization processing program, and a computer-readable recording medium.

[0007]

In order to achieve the above object, a threshold value setting method in a binarization process according to claim 1 of the present invention is such that a predetermined threshold value is set for an input image. The present invention relates to a method for setting a threshold value when performing binarization processing based on the above. The threshold value setting method in the binarization process is to set the threshold value based on the feature amount of the density histogram of the input image so that the threshold value dynamically follows the fluctuation of the input image. It is possible to

According to a second aspect of the present invention, in the threshold value setting method in the binarization process, in addition to the feature described in the first aspect, the feature amount is a peak included in the density histogram. It is characterized by being represented by a peak number in the inside and an offset amount from the peak position of the peak number.

Further, in the threshold value setting method in the binarization processing according to claim 3 of the present invention, in addition to the feature described in claim 1 or 2, the data regarding the input image is displayed as an image. A display unit and an input unit that allows the user to specify an arbitrary position for the image displayed on the display unit. Based on the information at the position specified by the user on the image, the threshold value is set. It is characterized by setting a value.

The user can also specify the threshold value from the density histogram by the peak position and the offset amount. In this method, a density histogram relating to the input image data is displayed, and a user specifies an arbitrary point on the density histogram waveform from an input unit, and based on the peak number and the offset amount of the specified point. The threshold value is set. As the input unit, various pointing devices such as a mouse, a tablet, a digitizer, a light pen, a keyboard, a numeric keypad, a jog dial, and a touch pad can be used. Further, the user can directly input a threshold value by a numerical value, or can select and specify it from options.

Furthermore, the threshold value setting method in the binarization processing according to claim 4 of the present invention is characterized in that, in addition to the feature described in any one of claims 1 to 3, a plurality of users are used. It is characterized by setting a threshold value.

Furthermore, in the threshold value setting method in the binarization processing according to claim 5 of the present invention, in addition to the feature described in any one of claims 1 to 4, The density threshold value is set based on the density value at the position designated by.

Further, in the threshold value setting method in the binarization processing according to claim 6 of the present invention, in addition to the feature described in any one of claims 1 to 5, The setting parameter value in the binarization process is set based on the density value of the position designated in step.

Furthermore, a binarization processing circuit according to a seventh aspect of the present invention is a circuit which performs binarization processing on an input image based on a predetermined threshold value. The binarization circuit includes an input image storage unit 1 that inputs and holds an input image from the outside, and a binarization processing unit that is connected to the input image storage unit 1 to read the input image and perform the binarization process. 3, a control unit 4 that creates a density histogram for performing the binarization processing in the binarization processing unit 3 and calculates a necessary threshold value, and a display unit that displays data regarding the input image as an image. 5 and an input unit 6 that allows the user to specify an arbitrary position for the image displayed on the display unit 5, and the user can use the input unit 6 to display the display unit 5
Based on the information that the position specified on the image has,
It is characterized by setting a threshold value.

Furthermore, a binarization processing program according to claim 8 of the present invention relates to a program for binarizing the input image based on a predetermined threshold value. The binarization processing program includes a unit for displaying data relating to the input image as an image, a unit for creating a density histogram based on the input image, and a peak number in a peak included in the density histogram of the input image. A unit for calculating a density threshold value based on a feature amount represented by an offset amount from the peak position of the peak number; a unit for the user to designate an arbitrary position for the displayed image; It is characterized by further comprising means for setting a density threshold value based on the characteristic amount obtained from the density value at the position specified above.
This program also includes a form that can be downloaded via a network.

Furthermore, a computer-readable recording medium according to claim 9 of the present invention is defined by claim 8.
The binarization processing program described in 1. is stored. The recording medium includes a CD-ROM, a CD-R, and a CD.
-RW, flexible disk, cassette tape, M
O, DVD-ROM, DVD-RAM, DVD-RW,
It includes a magnetic disk such as a DVD + RW, an optical disk, a magneto-optical disk, a semiconductor memory and other media capable of storing programs.

The threshold value setting method, the binarization processing circuit, the binarization processing program, and the computer-readable recording medium of the binarization processing of the present invention use the density histogram to set the threshold value. ing. The characteristics of the density histogram, that is, the positions where the peaks and valleys of the waveform appear do not change much even if there is illumination or ambient light. Therefore, the peak of the density histogram, that is, the peak, is captured as a feature, and the peak distance and the offset amount from that peak are used to identify how far away the arbitrary density specified by the user is from which peak, and based on the result. To determine the density threshold. By this method,
It is possible to dynamically determine even a plurality of arbitrary concentration threshold values.

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION Embodiments of the present invention will be described below with reference to the drawings. However, the following embodiments are readable by a threshold value setting method, a binarization processing circuit, a binarization processing program, and a computer in the binarization processing for embodying the technical idea of the present invention. The present invention exemplifies a recording medium, and the present invention does not specify a threshold value setting method, a binarization processing circuit, a binarization processing program, and a computer-readable recording medium in the following binarization processing. .

Further, in this specification, in order to facilitate understanding of the claims, the numbers corresponding to the members shown in the embodiments are referred to as "claims column" and "to solve the problems. It is added to the members shown in the column of "means". However, the members shown in the claims are not limited to the members of the embodiment. The sizes and positional relationships of members shown in the drawings may be exaggerated for clarity of explanation.

In the present specification, the threshold value setting method and the binarization processing circuit in the binarization process relate to the threshold value setting in the binarization process, the binarization process itself, and the binarization process. The present invention is not limited to an apparatus and method for performing processing by hardware, and is intended to include an apparatus or apparatus for performing software processing. For example, a general-purpose or dedicated computer, workstation, or terminal in which software, programs, plug-ins, objects, libraries, applets, compilers, etc. can be installed in a general-purpose circuit or computer to set threshold values for binarization processing. , Portable electronic devices,
PDA, pager, smart phone and other electronic devices are also included in the threshold value setting method and the binarization processing circuit in the binarization processing of the present invention. Further, in the present specification,
The program itself is also included in the binarization processing circuit.

In the present specification, the peak value refers to the concentration value at the position of the steeple having the maximum in each of the peaks of the waveform of the concentration histogram.

The binarization process is a process which is carried out as a pre-process of image processing or the like and which replaces a multi-valued image with a binary image above and below the specific threshold value. FIG. 1 shows an example of setting a threshold value in the density histogram of the input image. In a normal binarization process, since the threshold value is fixed, if a change occurs in the input multivalued image due to the influence of disturbance, a desired binary image may not be obtained. Therefore, a method is known in which the threshold value is not fixed and is varied so that the threshold value is always the optimum value according to the input image. This method is generically called an automatic threshold value determining method.

When the threshold value is determined, binarization processing is performed based on this value. Normally, once the threshold value is determined, this value is fixed and binarized. Therefore, when ambient light is added to the input image or there is a change such as illumination change, the brightness and density will change even if the object of the image itself is the same. It may not be optimal. Therefore, the desired result cannot be obtained even if the binarization process is performed with the same threshold value.

However, even when the input image changes in this way, when the density histograms are compared before and after the change, the waveform itself does not change much and maintains a constant form. 2A to 2D show density histograms of the input image data when the ambient brightness is changed for the same object. As shown in this figure, although the peak appearance position and height of the density histogram have changed, the number of peaks, the peak and valley patterns of the corresponding peaks, the mutual positional relationship, and the peak height relationship It can be seen that the shape of the waveform is almost maintained. Therefore, even if the input image changes from FIG. 1A to FIG. 1B as shown in FIG. 1, the threshold value is recalculated by designating the threshold value from the characteristic amount of the density histogram. The threshold can also be changed. In other words, if the threshold value is not fixed at a fixed value, but is specified as the feature amount of the density histogram, for example, the position of a mountain or a valley, specifically, the peak number and the offset amount, the change in the density or brightness of the image The density threshold value can be dynamically determined by following the above, and the optimum binarization processing can be maintained, so that the image processing is stabilized.

[Binarization Processing Circuit] The configuration of the binarization processing circuit according to the embodiment of the present invention will be described with reference to FIG. Figure 6
The binarization processing circuit shown in FIG. 1 is an input image storage unit 1 that inputs and holds an input image from the outside, and a binarization processing unit that is connected to the input image storage unit 1 to read the input image and perform binarization processing. 3, a control unit 4 that creates a density histogram for performing binarization processing in the binarization processing unit 3 and calculates a necessary threshold value, and a display unit 5 that displays data related to an input image as an image. The input unit 6 that allows the user to specify an arbitrary position for the image displayed on the display unit 5 is provided.
The relationship between the members and the connection form are not limited to the example shown in FIG. 6, and the binarization processing unit 3 and the control unit 4 may be replaced with each other or may be configured by one member.

The input image storage unit 1 stores the image input from the image source 7 connected thereto as data. For example, a real-time image taken by an imaging camera, a pre-recorded reproduced image, or the like is used as the input image.
The input image storage unit 1 uses a memory that stores the input image, for example, a RAM that is a volatile memory.

The binarization processing unit 3 serves as a work area for reading the input image from the input image storage unit 1 and performing the binarization process. The control unit 4 connected to the binarization processing unit 3 uses C
It is composed of a PU or the like, and executes the generation of a density histogram of an input image and its differentiation, the setting and calculation of a threshold value necessary for the binarization processing, the calculation of setting parameters, and the like. The setting parameter values include density thresholds such as upper and lower thresholds and frequency cut levels. Further, the threshold value setting method, the binarization processing circuit, the binarization processing program, and the computer-readable recording medium in the binarization processing of the present invention are not limited to the method of binarizing an input image, It can also be used in a method of obtaining grayscale data that has been converted into gradation, and is within the scope of the present invention regardless of the title of the invention. Data relating to various setting parameters and density histogram calculated according to a predetermined procedure are held in a memory inside the control unit 4.

The display unit 6 displays the data relating to the input image as an image. The image of the input image, the density histogram, or the differential waveform thereof can be sent from the control unit 4 to the display unit 5 as necessary and can be displayed on the preview screen provided in the display unit 5. A plurality of preview screens may be provided to separately display these images and waveforms, or they may be switched and displayed on one screen. Further, it is possible to display a point or a region which becomes a threshold on the displayed image or density histogram waveform.

Further, an input section 6 is provided which allows the user to designate an arbitrary position for the image displayed on the display section 5. The user can specify a desired portion for setting the density threshold value on the input unit 6 from the image displayed on the preview screen or the density histogram screen. As the input unit 6, various pointing devices such as a mouse, a tablet, a digitizer, a light pen, a keyboard, a numeric keypad, a jog dial, a touch pad, a slide pad, and AccuPoint can be used.

In order to set the threshold value, the user designates a desired position from the input section 6 on the preview screen on which the input image is displayed. Then, the position of the density value of the pixel at the designated position on the density histogram of the input image is obtained. The obtained position is specified by the peak number and the offset amount, and is displayed numerically as necessary and on the density histogram waveform. By this, the user can grasp the relationship between the density of the input image appearing on the screen using the actual image and the position in the density histogram. It can be intuitively set or selected.

Further, a plurality of threshold values can be set by this method. Conventionally, the threshold value can be calculated only one threshold value for one input image,
In addition, the threshold cannot be automatically set based on the part specified by the user. On the other hand, in the above method, the user can freely specify points on the screen, so that theoretically any number can be set. For example, in the example of FIG. 1, the upper limit threshold and the lower limit threshold are set, but other thresholds such as the frequency cut level and parameters used for the binarization processing can be determined. .

Generally, it is said that it is preferable to set the threshold value for the binarization process between the large peaks displayed in the density histogram. Therefore, a number is assigned to each peak and it is determined which threshold is set between which peak. The position to set the threshold,
That is, the density value set as the threshold value is expressed as an offset amount at the peak number and the position internally dividing the peak and the previous peak. Therefore, before setting the threshold value, the peak number N and the offset value P to be used as the threshold value are designated in advance. In this case, the threshold is N
Between the -1st peak and the Nth peak, the concentration value is a point at which these peak values are internally divided into P: (100-P). For example, when N = 3 and P = 67, these peak values are 67:33 between the second and third peak numbers.
The concentration that is internally divided into is the threshold value.

The procedure for executing the automatic threshold value determining method based on the threshold value will be described below with reference to the flowchart of FIG. 3 and the density histogram of FIG. In the following example, one threshold value is set for input image data of 255 gradations.

First, in step S1, input image data is fetched. Then, in step S2, a density histogram as shown in FIG. 4A is calculated based on the input image data. The density histogram (plHistgram) is 0 on the horizontal axis.
6 is a graph showing density values from (black) to 255 (white), and the vertical axis representing frequency, that is, the number of pixels included in input image data.

Next, in step S3, smoothing differentiation of the density histogram waveform is performed. The original waveform of the density histogram has a lot of jaggies due to noise contained in the original image, and many peaks may stand up as they are. Therefore, the smoothing process for smoothing the jaggies of the waveform is performed while performing the smoothing process with a predetermined smoothness. The smoothing range (dSmooth) can be performed over the entire region, or a specific range can be designated. As a result of performing the smoothing differentiation, the differential waveform as shown by the broken line in FIG. 4B is obtained as the smoothing component result (plDef).

When the noise setting cannot be removed completely, the lower limit threshold and the upper limit threshold can be set. By setting these thresholds, it is possible to exclude the concentrations lower than the lower limit threshold or higher than the upper limit threshold from the peak detection targets.

Further, a frequency cut level may be set. The frequency cut level is displayed as a ratio to all the pixel numbers, that is, the frequencies, and density values having a frequency lower than the value set at the frequency cut level are excluded. That is, the concentration having a low appearance ratio is excluded from the peak detection targets. For example, if the frequency cut level is set to 10%, the frequency represented by the vertical axis of the density histogram is 10% of the total frequency.
Concentrations below this will be excluded from the processing target.

Then, in step S4, the peak of the density histogram is detected. The position where the smoothed differential waveform of the density histogram crosses zero corresponds to a peak or a valley of the original density histogram waveform. Of these, a zero cross point corresponding to a mountain, that is, a peak is extracted. That is, FIG.
As shown in (c) and FIG. 5, the point where the smoothed differential result changes from positive to negative is the peak position of the density histogram. Figure 5
In the example of the density histogram shown in FIG.
170 and 220 are detected as peaks.

Then, an array of the concentration values of the detected peaks is obtained. This sequence is called the peak concentration sequence (plPLevel).
The peak concentration array has 256 arrays beforehand, and pl
PLevel [0], plPLevel [1], plPLevel [2], ..., plPLe
It is written as vel [255]. That is, up to 256 peaks can be detected. In this peak concentration array, the concentration values detected as peaks are stored in order from plPLevel [0]. The peaks are detected from the high density side (density 255 = white) to the low density side (density 0 = black). The start and end points of the detection range are also considered as peaks. As a result, plPLevel [0] = 255, plPLevel [1] = 220, plPL
evel [2] = 170, plPLevel [3] = 120, plPLevel [4]
= 41, plPLevel [5] = 0.

Further, in step S5, the threshold value is determined. Based on the peak number N and the offset value P for specifying the preset threshold value, FIG.
As shown in (c), the density value at the point where the density range between the Nth and N + 1th positions of the peak density array is internally divided into P: (100-P) is set as the density threshold value.

Then, the input image data is binarized by the density threshold value set in step S6. The above is the flow of the binarization process, and the binarization process is executed based on the set density threshold value.

[0042]

As described above, the threshold value setting method, the binarization processing circuit, the binarization processing program, and the computer-readable recording medium in the binarization processing of the present invention are the density of the input image. By setting the threshold value based on the feature amount of the histogram, so that an appropriate binarization processing result can be obtained even if the input image changes,
It is possible to dynamically adjust the threshold value and set an optimum density threshold value. For example, if the condition for determining the threshold value is specified by the peak number of the density histogram and the offset amount, the density threshold value is recalculated based on the set condition even if the input image changes and the density histogram changes. As a result, a new density threshold value is determined. In this way, it is possible to change to an appropriate threshold value by following changes in the input image, and a feature that can automatically determine the optimum threshold value according to the brightness of the input image and surrounding conditions is realized. It Since the user can directly specify the threshold value on the preview screen, the user can intuitively specify the threshold value, and there is also an advantage that a plurality of threshold values can be set.

[Brief description of drawings]

FIG. 1 is a graph showing a state in which an upper threshold value and a lower threshold value are set for a density histogram of an input image.

FIG. 2 is a graph showing a density histogram of input image data when ambient brightness is changed for the same object.

FIG. 3 is a flowchart showing a procedure for executing an automatic threshold value determining method based on a density threshold value.

FIG. 4 is a graph showing how a density threshold is set in a density histogram by automatic threshold determination.

FIG. 5 is a graph showing a density histogram of an input image and a result of smoothing and differentiating the same.

FIG. 6 is a block diagram showing a configuration of a binarization processing circuit according to an embodiment of the present invention.

[Explanation of symbols]

1. Input image storage unit 3 ... Binarization processing unit 4 ... Control unit 5 ... Display 6 ... Input section 7 ... Image source

   ─────────────────────────────────────────────────── ─── Continued front page    F-term (reference) 5B047 AA01 AB02 DB04 DC04                 5B057 BA24 BA29 BA30 CA08 CB06                       CE12 DA17 DB09 DC23                 5C077 LL08 LL09 LL10 PP02 PP47                       PQ12 PQ19 RR03 RR15 SS01                 5L096 AA06 CA24 FA37 GA51

Claims (9)

[Claims] 1. A method of setting a threshold value when performing binarization processing on an input image based on a predetermined threshold value, wherein the threshold value is set based on a feature amount of a density histogram of the input image. Is set, the threshold value can be dynamically tracked with respect to the fluctuation of the input image. 2. The binarization process according to claim 1, wherein the feature amount is represented by a peak number in a peak included in a density histogram and an offset amount from the peak position of the peak number. Threshold setting method. 3. A display unit for displaying data relating to the input image as an image, and an input unit capable of allowing a user to designate an arbitrary position with respect to the image displayed on the display unit. 3. The threshold value setting method in the binarization process according to claim 1, wherein the threshold value is set based on the information of the position specified above. 4. The threshold value setting method in the binarization process according to claim 1, wherein the user sets a plurality of threshold values. 5. The binarization process according to claim 1, wherein the density threshold value is set based on a density value at a position designated by the user on the image. Threshold setting method. 6. The binary value according to claim 1, wherein a setting parameter value in the binarization processing is set based on a density value at a position designated by the user on the image. Threshold setting method for sizing. 7. A binarization processing circuit for binarizing an input image based on a predetermined threshold value. An input image storage unit (1) for inputting and holding an input image from the outside.
And a binarization processing unit (3) that is connected to the input image storage unit (1) and reads the input image, and performs the binarization process, and the binarization processing by the binarization processing unit (3). A control unit (4) that creates a density histogram to perform processing and calculates a necessary threshold value, a display unit (5) that displays data related to the input image as an image, and a display unit (5) that displays the data. The input image (6) that allows the user to specify an arbitrary position for the displayed image, and the information that the position specified by the user on the image of the display part (5) by the user (6) has. A binarization processing circuit which sets a threshold value based on the above. 8. A binarization processing program for performing binarization processing on an input image based on a predetermined threshold value, means for displaying data relating to the input image as an image, and based on the input image Means for creating a density histogram; means for calculating a density threshold value based on a feature number represented by a peak number in a peak included in the density histogram of the input image and an offset amount from the peak position of the peak number A means for the user to specify an arbitrary position on the displayed image, and a means for setting a density threshold value based on the feature amount obtained from the density value at the position specified by the user on the image, A binarization processing program comprising: 9. A computer-readable recording medium storing the binarization processing program according to claim 8.