JPS59114687A - Binary coding method of picture data - Google Patents

Abstract

PURPOSE:To binary code a desired object and to discriminate the binary coded object from its periphery by setting a partial picture at a proper size. CONSTITUTION:The average of the picture data within a partial picture PKl is set at mKl and used as the threshold value at the center position qKl=(xK, yl) of PKl. The threshold values at the points other than {qKl; 1<=K<=m, 1<=l<= n} are decided as follows. That is, an optional point (i, j) on a picture included in P' is enclosed by four points forming an apex of a rectangle among {qKl} 1<=K<=m and 1<=l<=n. The threshold values ti and j at the point (i, j) are defined by an interpolation equation. A point (i', j') of P' closest to the point (i, j) is selected for the point (i, j) belonging to P'', and the threshold values ti' and j' of the point (i', j') obtained by an interpolation equation are defined as the threshold value tij of the (i, j). These threshold values thus obtained vary consecutively on a picture. The picture data value at the point (i, j) is defined as dij to obtain the binary picture data bij.

Description

[Detailed Description of the Invention] Industrial Application Field The present invention relates to image gradation data (hereinafter abbreviated as image data).
is divided into partial images, and the threshold value at each point is determined by an interpolation formula from the average value of the image data of each partial image.
The present invention relates to a method for binarizing image data.

The structure of the conventional example and its problems The image binarization method in the conventional example will be described below. Image data is represented by an array 1di, +1≦i<A, 1≦j≦8. A and B are image sizes in the X and Y coordinate directions, respectively. Initial threshold value t.

and 0(k(1), giving the coefficient k, the following recurrence formula (1)
and (2), the threshold value at each point +t, , +13
i<A, '+5; Calculate j≦B.

Image data (d・) and threshold value (from t, l, (3) Equation 1), binary image data ib,,+1iiA,1<j
Seek the hawk.

Incidentally, the horizontal image size is B140, and the initial value of the threshold value is t. FIG. 1 shows an example in which the threshold value for each point was determined by setting the coefficient k to 10 and the coefficient k to 0.2.

1 in FIG. 1 is image data on one horizontal line, and 2 in FIG. 1 is binary image data obtained from the threshold value on the same line obtained by the above method.

The disadvantage of this method is that changes in horizontal image data affect the threshold value, but changes in vertical image data do not affect the threshold value. There may be cases where it is not possible to do so.

OBJECTS OF THE INVENTION In view of the above-mentioned drawbacks, the present invention provides a method for binarizing image data, which captures and binarizes changes in brightness in the vertical direction in the same way as in the horizontal direction.

Structure of the Invention The present invention divides an image into m×n partial images of the same size by n horizontal lines and m vertical lines,
The average of the grayscale data of a partial image is set as the binarization threshold at the center point of the partial image, the threshold value at the m x n center points of the m x n partial images is determined, and then among those points, Adjacent four points located at the vertices of the rectangle 01 * fI L(tl, r2L(N2.JI
Binarization threshold t11J determined by the above method at L(+2.J2) (tl<12++1<+2)
1, 1 mountain 2, 112j1, t12j2, threshold value ti+ at any point (i, i) inside this rectangle is calculated using the interpolation formula % formula % ) (1) By obtaining a binarization threshold value and then comparing the gradation data value at each point on the image with the above binarization threshold value, you can create binary image data. Binarize and
Make it easier to distinguish the object from its surroundings.

DESCRIPTION OF EMBODIMENTS An image binarization method according to an embodiment of the present invention will be explained below with reference to the drawings. FIG. 3 is an image of AxBO size. This is divided into m x n partial images Pkl of the same size. (1≦≦m, 1≦l≦n) FIG. 4 is a distribution diagram of the center position qkl=(xk, yl) tv of each partial image Pkl. Let the average f rnk,g of the image data in Pkl be the center position of P kit qk1= (xk,
Let the threshold value tXk at yl) be 7i. Iqk
The threshold values at points other than l; 1≦k<m, 1≦l≦n are determined as follows.

Figure 5 shows the entire image as ql 11qm1 +qmn+
A diagram showing the rectangular y area formed by sequentially connecting the four points of q1n and the area p // outside of the rectangular area. An arbitrary point of the image-F that is included in P' (i, D are 'qM' 1 < k
<m, 1<71<B, a certain 4 that forms the vertex of the rectangle
surrounded by points. That is, X k'-1 < X k+ s,
, 71≦j≦”7+1, the four points surrounding N1j) where 7 exists are (xl(+3'k) r
(xlc+Vk-1-1L("k+1 tVk)l (
The threshold value yl at (i+j)K is determined by the following interpolation formula (2).

tij””! ”k+1 1)(y7+1) doortxk
yf("k+1-ri(j-y7) ・txk74(,+
(11()(y,,-2)・txh+yp<'xh)
(j-5r7)・tX1+t, 7. g+t+/”k
+1-Xk) (y, +1-yl) =-Mountain--
Old = - (5) (Equation (5) is equation (1), xk = 111xk
+1=12, yl=i113'l+1="2.) For the point (i, i) belonging to P", the point of P' closest to (i, j) (z/, j/ ), and the threshold value 1. of (i'+]') obtained by interpolation formula (2). ．． -/(iI
+) Threshold "+]" is set to a value t.The threshold value obtained above changes continuously on the image 1].The image data value at point (i, i) is divided by d... , (6), the binary image data 1 is determined.

In addition to the effects of the invention, according to the present invention, it is possible to binarize the target object by appropriately setting the partial image size even for images with small differences in brightness and uneven illumination. It became easier to distinguish it from its surroundings.

[Brief explanation of the drawing]

Figure 1 is a graph of image data and threshold values obtained using the conventional method, and Figure 2 is a graph of threshold values obtained from the data in Figure 1.
Graph of value image data, Figure 3 is an explanatory diagram of a method of decomposing an image into partial images, Figure 4 is a distribution diagram of the center position of each partial image, and Figure 5 is a binarization diagram of an embodiment of the present invention. It is an explanatory diagram of a method. Name of agent: Patent attorney Toshio Nakao and 1 other person No. 1
figure

Claims (1)

[Claims] An image is divided into m by n horizontal lines and m vertical lines.
Divide into X n partial images of the same size, set the average of the grayscale data of the partial images as the binarization threshold at the center point of the partial image, and set m of the m x n partial images.
x Define the threshold values at n center points, and then calculate the four adjacent points located at the vertices of the rectangle among those points ('1+] IL (11+] 2) D2+ 11 L
Binarization threshold value ti1 j determined by the above method in (12+12) ('1<12t 31<12)
From l Iti1j2' t2j11ti2j2, threshold value ti at any point (i, i) inside this rectangle, 3 k interpolation formula % Formula % (1) Calculate the binarization threshold value at each point on the image. A method for binarizing image data, in which binary image data is created by obtaining the gradation data value at each point on the image and then comparing the binarization threshold with the binarization threshold.