JPH0652304A - Device and method for processing image - Google Patents

Abstract

PURPOSE:To reduce the burden of following processing as much as possible in the case of dispatching the contour point of a provided binary image to the following processing. CONSTITUTION:The binary image provided from a binary image possessing means 11 is extracted by an outline vector extracting means 12 and smoothed by eliminating any unwanted vector at an outline vector detecting, reducing, smoothing and magnifying means 13. Afterwards, a binary image is generated by a binary image reproducing means 14 based on the outline vector processed by the outline vector detecting, reducing, smoothing and magnifying means 13, and outputted by a binary image output means 15.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing apparatus for extracting contour information from a binary image and subjecting it to predetermined processing.

[0002]

2. Description of the Related Art As an apparatus of this type, the applicant of the present application has already filed Japanese Patent Application No. 3-345062 or Japanese Patent Application No. 4-169.
Submitted as No. 581.

In all of these proposals, when a binary image is scaled and output, the binary image itself is not scaled, but the contour information of the binary image is extracted and the extracted contour information is extracted. This is done in order to obtain a high quality image by generating a scaled image based on.

Specifically, Japanese Patent Application No. 3-345062 discloses an outline vector extracted from a binary image and smoothly scaled at a desired magnification (arbitrary) in the state of the extracted outline vector expression. A vector is created and a binary image is regenerated from the smoothly scaled outline vector. By this, it is intended to obtain a high-quality digital binary image which is scaled at a desired magnification (arbitrary).

The main parts will be outlined below. Figure 10
FIG. 4 is a diagram best representing the features disclosed in Japanese Patent Application No. 3-345062.

In FIG. 1, reference numeral 1 denotes a binary image acquisition means for acquiring a digital binary image to be scaled and outputting a binary image in a raster scanning format. Reference numeral 2 denotes a rough contour from a binary image in a raster scanning format. An outline extracting means for extracting a vector (an outline vector before performing the smoothing / scaling processing), 3 is an outline smoothing / scaling means for smoothing and scaling the rough contour vector data in the form of vector data. Binary image reproducing means for reproducing binary image data in raster scanning format from outline vector data, 5
Is a binary image output means for displaying raster scanning type binary image data, taking a hard copy, or outputting to a communication path or the like.

The binary image acquisition means 1 is composed of, for example, a known raster scanning type binary image output device which reads an original image as a binary image and outputs it in a raster scanning format. The outline extracting means 2 is composed of, for example, the device described in Japanese Patent Application No. 2-281958 previously proposed by the applicant of the present application. FIG. 2 shows the scanning form of the raster scanning type binary image data output from the binary image acquisition unit 1, and the scanning form of the raster scanning type binary image data input by the outline extracting unit 2. Also shows. In such a format, the outline extracting means 2 inputs the raster scanning type binary image data output by the binary image acquiring means 1. In FIG. 2, 101 is
It shows a pixel in a binary image during raster scanning, 1
Reference numeral 02 represents a 9-pixel area including 8 pixels in the vicinity of the pixel 101. As mentioned above, Japanese Patent Application No. 2-28195.
The outline extracting means described in No. 8 moves the pixel of interest in raster scanning order, and pays attention to each pixel of interest in accordance with the state (white pixel or black pixel) of each pixel in the 9 pixel region 102. The contour edge vector (horizontal vector or vertical vector) existing between the pixel and the neighboring pixel of the pixel of interest is detected, and if the contour edge vector exists, the starting point coordinate and orientation data of the edge vector are extracted. Then, the rough contour vector is extracted while updating the connection relationship between the edge vectors.

FIG. 3 shows an example of the extraction state of the contour side vector between the target pixel and the neighboring pixels of the target pixel. In the figure, the Δ mark represents the starting point of the vertical vector, and the ∘ mark represents the starting point of the horizontal vector.

FIG. 4 shows an example of the rough contour vector loop extracted by the outline extracting means described above. Here, each square that is divided by a grid indicates a pixel position of the input image, a blank square means a white pixel, and a circle formed by a dot pattern means a black pixel. Similar to FIG. 3, the Δ mark represents the starting point of the vertical vector, and the ∘ mark represents the starting point of the horizontal vector.

As can be seen from the example of FIG. 4, the outline extraction means extracts a region where black pixels are connected as a rough contour vector loop in which horizontal vectors and vertical vectors alternate (always alternate). However, in this case, the direction of advancing the extraction process is such that the right side of the advancing direction is the black pixel region. The starting point of each rough contour vector is extracted as the intermediate position of each pixel of the input image. That is, when the existence position of each pixel is represented by an integer (x, y), the starting point of the extracted vector has a value obtained by adding 0.5 to each coordinate value or a value obtained by subtracting 0.5. More specifically, the one-pixel wide line portion in the original image is also extracted as a rough contour loop having a significant width.
The rough contour vector loop group extracted in this way has a data format as shown in FIG.
Will be output. That is, it is composed of the total number a of rough contour loops extracted from the image and each rough contour loop data group from the first contour loop to the a-th contour loop. Each rough contour loop data includes the total number of starting points of the contour side vectors existing in the rough contour loop (which can also be considered as the total number of contour side vectors) and the starting point coordinates of each contour side vector in the order of forming the loop. (X coordinate value, y coordinate value) value (start point of horizontal vector and start point of vertical vector are arranged alternately).

Now, in the outline smoothing / magnifying unit 3 shown in FIG. 10, the rough contour vector data (see FIG. 5) output from the outline extracting unit 2 is input, and the smoothing and the desired scaling factor are obtained. The scaling process is performed in the form of outline vector data (coordinate values). FIG. 6 shows a more detailed structure of the outline smoothing / magnifying unit. In FIG. 6, reference numeral 31 is a magnification setting means for changing magnification, and 3
Reference numeral 2 is a first smoothing / magnifying unit. The first smoothing / magnifying unit smoothes and scales the input rough contour data at the magnification set by the magnification setting unit 31. The processing result is further smoothed by the second smoothing means 33 to obtain a final output.

The magnification setting means 31 may pass a value preset by a dip switch, a dial switch or the like to the first smoothing / magnifying means, or an I / F (interface) from some outside. ), And means for giving information on how many times to multiply the image size given as input independently in the main scanning (horizontal) direction and the sub-scanning (vertical) direction. Is.

The first smoothing / magnifying means 32 obtains magnification information from the magnification setting means 31 and performs smoothing / magnifying processing.

FIG. 7 shows an example of the hardware configuration for implementing the outline smoothing / magnifying unit. In FIG. 7, 71
Is a CPU, 72 is a disk device, and 73 is a disk I /
O and 74 are R storing the operation processing procedure of the CPU 71.
OM. 75 is an I / O port, 76 is a RAM (random access memory), and 77 is a bus connecting the above blocks.

The output of the outline extracting means shown in FIG. 2 is stored as a file (coarse contour vector data) in the disk device 72 in the data format shown in FIG.

The CPU 71 operates according to the procedure given in FIG. 8 and executes outline smoothing / magnifying processing.

First, in step S1, the disk I / O 73
The rough contour data stored in the disk device 72 is read out via the above and is read into a working memory area (not shown) in the RAM 76. Next, in step S2, the first smoothing and scaling processing is performed.

The first smoothing process is performed for each closed loop unit of the rough contour data. Focusing on each contour edge (horizontal vector or vertical vector) vector of each rough contour data in sequence, up to three vectors before and after each contour edge vector (that is, before the attention edge
The book, the side of interest itself, and a total of 7 after the side of interest and a total of 7
The pattern is divided according to the combination of the length and the direction of the side vectors that are continuous to each other (up to the side vector up to the book), and in each case, the contour points after the first smoothing that are the first smoothing results for the target side are obtained. I will define it. Then, the coordinate value of the contour point after the first smoothing and additional information indicating whether or not the contour point is a corner point (hereinafter referred to as corner point information) are output. The corner point here means a point located at a meaningful corner, and a corner point due to a notch or a jagged portion due to noise or other factors is excluded. Now, the contour points after the first smoothing determined as corner points (hereinafter referred to as corner points)
Is treated as a fixed point, that is, its position which is not smoothed by the second smoothing performed later. Further, the contour points after the first smoothing which are not determined as the corner points (hereinafter referred to as non-corner points) are further smoothed by the second smoothing later.

FIG. 9 shows this state, that is, the target rough contour side vector D _i , the three side vectors D _i-1 , D _i-2 and D _i-3 before the target rough contour side vector and the target. State of three edge vectors D _{i + 1} , D _{i + 2} , D _{i + 3} after the rough contour edge vector, and appearance of the contour points after the first smoothing defined for the target edge D _i Is shown.

The details of the first smoothing process have been described above.
The data after the first smoothing is sequentially constructed on a predetermined area of the RAM 76. Thus, after finishing the process of step S2 in FIG. 8, the CPU 72 performs the second smoothing process of step S3.

In the second smoothing, the data after the first smoothing is input and processed. That is, the number of closed loops, the number of contour points for each closed loop, the coordinate value data string of the first smoothed contour points for each closed loop, and the additional information data string of the first smoothed contour points for each closed loop. Input and output the contour point data after the second smoothing.

As shown in FIG. 11, the contour data after the second smoothing is composed of a closed loop number, a contour point number table for each closed loop, and a coordinate value data string of the second smoothed contour point for each closed loop. To be done.

The outline of the second smoothing process will be described below with reference to FIG. Similar to the first smoothing, the second smoothing is processed for each contour loop unit, and the processing is advanced for each contour point in each contour loop.

For each contour point, if the contour point of interest is a corner point, the input contour point coordinate value itself is
The second smoothed contour point coordinate data for the target contour point is set. That is, nothing is changed.

If the contour point of interest is a non-corner point, the coordinate value obtained by weighted averaging of the coordinate value of the front and rear contour points and the coordinate value of the contour point of interest is used. The second smoothed contour point coordinate value for the point is used. That is, a target input contour point that is a non-corner point is defined as P _i (x _i , y _i ).
And the immediately preceding contour point in the input contour loop of Pi is P
_{i −1} (x _{i + 1} , y _{i + 1} ), the immediately following contour point is P _{i + 1} (x
_{i + 1} , y _{i + 1} ), and Q _i (x ′ _i , y ′ _i ) the second smoothed contour point for the target input contour point P _i , x ′ _i = k _i- Calculate as ₁ · x _i-1 + k _i · x _i + k _{i + 1} · x _{i + 1} y ′ _i = k _i-1 · y _i-1 + k _i · y _i + k _{i + 1} · y _{i + 1} . Here, k _i-1 = k _{i + 1} = 1/4, k
_i = 1/2.

In FIG. 12, point P₀ , P₁ , P₂ , P
₃ , P_Four Is the input, the first smoothed continuous contour point
Is part of a row, P₀ And P_Four Is the corner point, P₁, P₂ Over
And P₃Indicates a non-corner point. The processing result at this time is
Each point Q₀ , Q₁, Q₂ , Q ₃ , Q_Four Indicated by
It P₀ And P_Four Are angular points, so their coordinate values
But as it is Q respectively₀ And Q_Four Is the coordinate value of. Well
Point Q₁ Is P₀ , P₁ , P ₂ From the above equation
It has the calculated values as coordinate values. Similarly, Q₂ Is P₁，
P₂ , P₃ From Q₃ Is P₂ , P₃ , P_Four According to the above formula
The calculated value is used as the coordinate value.

The CPU 71 executes the above processing in the RAM.
A second smoothing process is performed on the first smoothed contour data in a predetermined area on 76. This process proceeds from the first loop to the second loop and the third loop in sequence for each loop, and the process for all the loops ends, thereby ending the second smoothing process. In the processing of each loop, the processing proceeds from the first point to the second point and the third point in order, and when the processing shown in the equation for all the contour points in the loop is completed, the processing of the loop is performed. After that, the process goes to the next loop.

When there are L contour points in the loop, the point before the first point is the Lth point, and the Lth point.
The point after the point is the first point. As described above, in the second smoothing, the same number of total loops as the first smoothed contour data to be input is used, and the number of contour points on each loop does not change, and the same number of contour point data is generated. The CPU 72
The above result is output to another area of the RAM 76 or to the disk device 72 in the form shown in FIG. 11, and the process of the second smoothing process (step S3) ends.

Next, the CPU 71 proceeds to step S4,
The data obtained as a result of the second smoothing is transferred to the binary image reproducing means 4 via the I / O 75, and the series of processing shown in FIG. 8 is completed.

The binary image reproducing means 4 can be constituted by, for example, the device described in Japanese Patent Application No. 3-172098 previously proposed by the present applicant. According to the apparatus, a binary image generated by filling the area surrounded by the vector graphic represented by the contour data based on the second smoothed contour data transferred via the I / O. Can be output in a raster scan format. Further, the proposal is to visualize using a binary image output means such as a video printer as described in the description.

The proposal of Japanese Patent Application No. 4-169581 is a further improvement of the above-mentioned Japanese Patent Application No. 3-345062, in which a low-magnification image does not become fat. Is. That is, Japanese Patent Application No. 3-3450
In the outline extraction unit of No. 62, the boundary between the white pixels and the black pixels of the original image is exactly the vector extraction target, whereas in this proposal, the black pixels between the black pixels are closer to each other (the black pixel area is Extract narrower than the area),
In addition, it is modified to perform outline smoothing in accordance with this.

[0032]

In the above-mentioned conventional example, since the contour points are extracted and defined for each pixel of the original image,
A binary image with sufficiently high image quality can be obtained by expressing all contour points in a short vector format with straight lines, but in the case of a complicated original image with many diagonal lines,
It turns out that there is a problem that the number of outline vectors becomes very large. The increase in the amount of data means that a large memory capacity to be consumed is required, and that the processing time required for smoothing and reproduction also increases.

[0033]

SUMMARY OF THE INVENTION The present invention has been made in view of the above problems, and when the contour points of the obtained binary image are passed to the subsequent processing, the load of the subsequent processing is reduced as much as possible. (EN) An image processing apparatus and method for enabling the above.

In order to solve this problem, the image processing apparatus of the present invention has the following configuration. That is, based on the contour vector acquisition means for acquiring the contour vector data extracted from the binary image and the inclination of a predetermined number of vector groups adjacent to each other acquired by the contour vector acquisition means, A detection unit that detects an unnecessary vector and a deletion unit that deletes the vector detected by the detection unit are provided.

The image processing method of the present invention comprises the following steps. That is, a process of acquiring contour vector data extracted from a binary image, and a process of detecting an unnecessary vector from the vector group based on the inclinations of the acquired predetermined number of vector groups, A deletion step of deleting the detected vector.

[0036]

In the structure and process of the present invention, in the obtained contour vector, an unnecessary vector is detected from the vector group based on a predetermined number of continuous vector groups, and the detected vector is deleted. .

[0037]

Embodiments of the present invention will be described in detail below with reference to the accompanying drawings.

FIG. 1 is a block diagram of the image processing apparatus of the embodiment. FIG. 10 previously proposed by the applicant of the present application
The difference from the block diagram of FIG. 3 is that the outline smoothing / magnifying unit 3 detects outline vector detection / deletion / smoothing /
This is that it is configured as the variable power unit 13. Other binary image acquisition means 11, outline extraction means 12, 2
The value image reproducing means 14 and the binary image output means 15 can be configured similarly to the reference numerals 1, 2, 4 and 5 in FIG. For this reason, the description of these is omitted to avoid duplication, and the outline vector detection / deletion / smoothing / magnifying unit 13 of FIG. 1 will be described below.

A more detailed structure of the outline vector detecting / reducing / smoothing / magnifying unit 13 in the embodiment is shown in FIG.
3 shows. In the figure, reference numeral 21 is a magnification setting means for variable magnification.
Reference numeral 22 denotes a first smoothing and scaling unit that smoothes and scales the input rough contour data at the scaling factor set by the scaling factor setting unit 21. The processing result is passed to the continuous vector detecting / deleting means 24 of the same direction, and continuous short vectors having the same slope are detected. If there are these short vectors, these are connected and the number of vectors to be processed thereafter is calculated. To reduce. Same direction continuous vector detection
The processing result of the deleting means 24 is further smoothed by the second smoothing means 23 to be the final output.

The units denoted by reference numerals 21, 22, and 23 in FIG. 13 can be configured in the same manner as the reference numerals 31, 32, and 33 in the above-mentioned prior patent proposal (FIG. 6).
reference). Further, the outline vector detecting / reducing / smoothing / magnifying unit of FIG. 13 is processed by a program in the case of the embodiment, and thus can be basically realized by the configuration of FIG. However, the programs stored in the ROM 74 are naturally different.

Continuous vector detection / deletion means 24 in the same direction
The outline of the operation will be described with reference to FIG. In the figure,
P _k−2 , P _k−1 , P _k , P _{k + 1} , P _{k + 2} ... Are mutually on the same contour vector loop forming the first smoothed / scaled contour vector which is an input. It is a sequence of continuous contour points. Each contour point P _i (i = ... k-2, k-1, k, k + 1, k + 2,
When the coordinate value of () is expressed by (x _Pi , y _Pi ), a short vector v _Pi (Δx) defined by contour points Pi and Pi + 1
_Pi, △ y _Pi) can be represented by _{△ x Pi = x Pi + 1} -x Pi △ y Pi = y Pi + 1 -y Pi.

[0042] Now, the outline point sequence to each other continuously on the same outline vector loop P _i-1, P _i, between _{P i + 1, △ x Pi} -1 × m = △ x Pi and △ y _Pi-1 If xm = Δy _Pi (where m is a real constant) holds, v defined by P _i-1 and P _i
Pi-1 and v _p defined by P _i and P _{i + 1} have the same slope.

In FIG. 14, ... P _K-2 , P _k-1 , P _k , P
_In the contour point sequence of _{k + 1} , P _{k + 2,} ..., The target contour point Pk is deleted, and ..., P _k-2 , P _k-1 , P _{k + 1} , P _{k + 2} ... It is shown that they are respectively output as Q _j-2 , Q _j-1 , Q _j , Q _{k + 1} . That is, v _Pk-1 and v _Pk are Δx _Pk-1
= Δx _Pk and _{Δy Pk-1} = Δy _Pk , and it is possible to detect that they have the same inclination because the condition shown by the equation is satisfied. Therefore, they are connected and treated as v _Qj-1 , and the target contour point P _k between P _k-1 and P _{k + 1} is deleted and excluded from the output target.

FIG. 15 is the same as FIG. 14 ... M _k-2 , M _k-1 , M
_k , M _{k + 1} , M _{k + 2} , ... Consecutive contour point sequences on the same contour vector loop forming the first smoothed / scaled contour vector to be input. Again, remove the _{M k, ..., M k-} 2, M k-1, M k + 1, M k + 2, ... have, in Tame new, _{respectively, ... N j-2, N} j-1 , N _j , N _{j + 1} , ...

That is, v_Mk-1And v_MkAnd at x_Mk-1
× 1/3 = Δx_MkAnd △ y_Mk-1× 1/3 = △ y_MkAnd
Satisfies the condition shown by the formula. Therefore, the same
Detecting that it has a tilt, v_Mk-1And v_MkAnd
Tied v_Nj-1Treated as M _k-1And M_{k + 1}The ring between
Border point M_k Is deleted and output.

The principle of deleting one central point of the contour point sequence consisting of three consecutive points on the same contour vector loop has been described above.

Here, since the slopes of the two contour vectors defined by these three points are equal, one point at the center is deleted and
The slope is the same as the two vectors before deletion, and the length is 2
The difference in the processing results after that which occurs when the two vectors before deletion are replaced with the contour vector equal to the sum of the lengths of the two vectors will be described.

That is, in the case where the same inclination contour vector is continuous, the following will consider how long it should be deleted.

In FIG. 16, ... a_k-2, a_k-1, a_k, a
_{k + 1}, a_{k + 2}, ... are contour vectors that have undergone the first smoothing and scaling processing
Are continuous with each other on the same contour vector loop
It is a contour point sequence to be performed. Where this ... a_k-2, a_k-1, a_k,
a_{k + 1}, a_{k + 2}When the second smoothing is performed as is, (...
Then a_k-2And a_{k + 2}Is the corner point, a_k-1, a_k, a_{k + 1}Is a non-corner point
And the obtained second smoothed contour point sequence
, ... b_k-2, b_k-1, b_k, b _{k + 1}, b_{k + 2}Expressed as ...
It On the other hand, ... c_j-2, c_j-1, c_j, c_{j + 1}, ... is 3 points, a_{k- 1},
a_k, a_{k + 1} Since the equation holds in_k
To delete, ..., a_k-2, a_k-1 , a_{k + 1} , a_{k + 2} Composed of,…
The sequence of continuous contour points
The contour point sequence after smoothing is ... C_j-2, c_j-1, c_j, c_{j + 1},…so
It is expressed.

In FIG. 17, ... D _k-2 , d _k-1 , d _k , d
_{k + 1} , d _{k + 2} , ... Are columns of another continuous contour point sequence on the same contour vector loop which constitutes the first smoothed / scaled contour vector. Where this ...
When the second smoothing is performed on d _k-2 , d _k-1 , d _k , d _{k + 1} , d _{k + 2} , ... (where d _k-2 and d _{k + 2} are corner points, and d _k-1 , d _k ,
It is assumed that d _{k + 1} is a non-angular point), and the obtained second smoothed contour point sequence is ... E _{k -2} , e _k-1 , e _k , e _{k + 1} , e _{k + 2} ...
It is expressed by. On the other hand, ... f _j-2 , f _j-1 , f _j , f _{j + 1} , ...
Since the equation is established at the three points d _k-1 , d _k , d _{k + 1} , d _k is deleted and ... D _k-2 , d _k-1 , d _{k + 1} , d
_{k + 2} , ... Consecutive continuous contour point sequences, and the second smoothed contour point sequence is ... F _j-2 , f _j-1 , f
It is represented by _j , f _{j + 1} , ....

In FIG. 16, b _k-2 , b _k-1 , b _k , b _{k + 1} , b _{k + 2} ,
, And ... C _j-2 , c _j-1 , c _j , c _{j + 1} , ..., The contour shapes represented are obviously different. Also, in FIG.
_k-2 , e _k-1 ,, e _k , e _{k + 1} , e _{k + 2} , ... And ... f _j-2 , f _j-1 ,, f _j ,
The contour shapes represented by f _{j + 1} , ... Are also obviously different.

This difference is shown in FIG.
As described in 2, the non-angular contour points take the weighted average described in the prior art between the immediately preceding contour point and the immediately following contour point on the same contour vector loop,
This is because the contour point immediately before or immediately after that is removed by the above-described contour point removal processing.

In FIG. 18, ... g_k-3, g_k-2, g_k-1, g
_k, g_{k + 1}, g_{k + 2}, g_{k + 3}, ... have undergone the first smoothing and scaling processing
Mutually on the same contour vector loop that makes up the contour vector
It is a continuous sequence of contour points. Where these ... g_k-3,
g_k-2, g_k-1, g_k, g_{k + 1}, g_{k + 2}, g_{k + 3}, ... as it is second
When smoothed (however, g_k-2And g_{k + 2}Is the corner point and g
_k-1, g_k, g_{k + 1}Is a contour point that is not a corner point)
The contour point sequence after the second smoothing obtained in_k-3, h
_k-2, h_k-1, h_k, h_{k + 1}, h_{k + 2}, h_{k + 3}It is expressed by ....
On the other hand, ... l_j-3, l_j-2, l_j-1, l_j, l_{j + 1}, l_{j + 2}, ... is 5
Point g_k-2, g_k-1, g_k, g_{k + 1}, g_{k + 2}Where g_k-2,
g_k-1, g_kAnd g_k-1, g_k, g_{k + 1}And g_k, g_{k + 1}, g_{k +2}of
The formula holds for all three consecutive points
G at the center when_k, And ... g
_k-3, g_k-2, g_k-1, g_{k + 1}, g_{k + 2}, g_{k + 3}, Composed of
To the continuous contour point sequence, and after the second smoothing for this
Represents the sequence of contour points.

FIG. 19 is a circuit diagram of FIG. 18 ... G _k-3 , g _k-2 ,
In g _k-1 , g _k , g _{k + 1} , g _{k + 3} , ..., g _k-2 and g _{k + 2} are also g
Similarly to _k-1 , g _k , and g _{k + 1} , the second smoothed contour point sequence obtained when the second smoothing is performed as it is is a contour point that is not a corner point ... h ′ _{k- 3} , h ' _k-2 , h' _k-1 , h ' _k ,
It is represented by _{h'k + 1} , _{h'k + 2} , _{h'k + 3} , ....

On the other hand, ... l' _j-3 , l' _j-2 , l' _j-1 , l' _j , l '
_{j + 1, l 'j +} 2, ... are 5-point _{g k-2, g k-} 1, g k, in _{g k + 1, g k +} 2, g k-2, g k-1, g _k and g _k-1 , g _k , g _{k + 1} and g _k ,
Since the formula is established for all the pairs of three consecutive points of g _{k + 1} and g _{k + 2} , the central g _k is deleted, and g _k-3 , g _k-2 , .., which are consecutive contour points formed by g _k-1 , g _{k + 1} , g _{k + 2} , g _{k + 3} , ...

Comparing FIG. 18 and FIG. 19, both are
(2) The state before the smoothing process is ..._k-3, g_k-2, g_k-1, g_k, g
_{k + 1}, g_{k + 2}, g_{k + 3}The coordinate values of, ... Are exactly the same, and in FIG.
g_{k- 2}And g_{k + 2}Is a corner point, whereas in FIG.
_k-2And g_{k + 2}Is different in that is a non-angular point
It g_k-2And g_{k + 2}In the case of FIG. 18, where is a corner point, g_k
Does not omit g_k-2, g_k-1, g_k, g_{k + 1}, g_{k + 2}Obtained from
The contour point sequence h after the second smoothing process_k-2, h_k-1, h_k, h_{k + 1},
h_{k + 2}The contour shape expressed by_kOmitting g _k-2, g
_k-1, g_{k + 1}, g_{k + 2}The second smoothed contour obtained by
Point sequence l_j-2, l_j-1, l_j, l_{j + 1}What is the contour shape represented by
It is a perfect match. On the other hand, g_k-2And g_{k + 2}Is a non-corner point
In the case of FIG._kDoes not omit g_k-2, g_k-1,
g_k, g_{k + 1}, g_{k +2}The second smoothed contour obtained by
Point sequence h '_k-2, h '_k-1, h '_k, h '_{k + 1}, h '_{k + 2}The ring represented by
Guo shape and g_kOmitting g_k-2, g_k-1, g_{k + 1}, g_{k + 2}Than
The obtained contour point sequence l ′ after the second smoothing processing_j-2, l '_j-1,
l '_j, l '_{j + 1}The contour shape expressed by is l '_j-1, l '_jAmong
The parts match exactly, but l '_j-2(= H '_k-2)When
l '_{j- 1}Between, and l '_jAnd l '_{j + 1}(= H '_{k + 2}Between)
Are not in agreement.

In FIG. 20, ... t_k-4, t_k-3, t_k-2, t
_k-1, t_k, t_{k + 1}, t_{k + 2}, t_{k + 3}, t_{k + 4}, ... is the first smoothing
Same contour vector that constitutes the scaled contour vector
It is a sequence of consecutive contour points on the loop. Where
Of ... t_k-4, t_k-3, t_k-2, t_{k- 1}, t_k, t_{k + 1}, t_{k + 2}, t_{k + 3},
t_{k + 4}When the second smoothing is performed as is (..., t
_k-3, t_k-2, ..., t_{k + 3}Are all non-angular points),
The obtained second smoothed contour point sequence is ... U_k-4, u_k-3, u
_k-2, u_k-1, u_k, u_{k + 1}, u_{k + 2}, u_{k + 3}, u_{k + 4}Express with…
is there. Also ... w_j-4, w_j-3, w_j-2, w_j-1, w_j, w_{j + 1}, w
_{j + 2}, w_{j + 3}, ... is 7 points t_k-3, t_k-2, t_k-1, t_k, t_{k + 1}, t
_{k + 2}, t_{k + 3}At t_k-3, t_k-2, t_k-1And t_k-2, t
_k-1, t_kAnd t_k-1, t_k, t_{k + 1}And t_k, t_{k + 1}, t_{k + 2}All of
The state where the formula is satisfied by a set of three consecutive points
, The central t_kIs deleted, and… t_k-4, t_k-3,
t_k-2, t_k-1, t_{k + 1}, t_{k + 2}, t_{k + 3}, t_{k + 4}Composed of, ...
Second smoothing for the continuous contour point sequence
The subsequent contour point sequence is represented. In this case u_k-3, u_k-2,
u_k-1, u_k, u_{k + 1}, u_{k + 2}, u_{k + 3}Contour shape represented by
And w_j-3, w_j-2, w_j-1, w_j, w_{j + 1}, w_{j + 2}Represented by
It perfectly matches the contour shape.

From the example described above, after the first smoothing,
A contour point z in the contour point sequence _kCentered on each other
7 consecutive contour points z_k-3, z_k-2, z_k-1, z_k, z
_{k + 1}, z_{k +2}, z_{k + 3}Defined between these 7 points in
If all 6 short vectors have the same slope
Is the contour point z_k, And the contour point sequence consisting of 6 points
z_k-3, z_k-2, z_k-1, z_{k + 1}, z_{k + 2}, z_{k + 3}After that
Even if processing is performed, the obtained contour shape does not change.
Can be easily understood. Also, z_k-2Or z_{k + 2}Is a corner point
Z, if any_k-2, z_k-1, z_k, z_{k + 1}, z_{k + 2}, z
_{k + 3}, Or z_k-3, z_{k -2}, z_k-1, z_k, z_{k + 1}, z_{k + 2}Z
_kBetween 6 points in a sequence of 6 consecutive contour points centered at
The slopes of all five short vectors defined by
Contour point z_kBy deleting the circle consisting of 5 points
Each boundary point sequence is z_k-2, z_k-1, z_{k + 1}, z_{k +2}, z_{k + 3}Well
Or z_k-3, z_k-2, z_k-1, z_{k + 1}, z_{k + 2}After that
Even if you do the reasoning, the obtained contour shape may not change
Easy to understand. z_{k + 2}And z_k-2Are both corner points
In the case of z_k-2, z_k-1, z_k, z_{k + 1}, z_{k + 2}Z_kCentered on
A series of 5 contour points that are continuous with each other
All four short vectors have the same slope
And the contour point z_kBy deleting the contour point sequence consisting of 4 points
z_k-2, z_k-1, z_{k + 1}, z_{k + 2}Perform the following processing with
It is also easy to understand that there is no change in the obtained contour shape.
Wear.

Next, the continuous vector detecting / deleting means 24 in the same direction in FIG. 13 will be described. The continuous vector detecting / deleting means 24 in the same direction can be realized by using the hardware shown in FIG. 7, and its operation will be described with reference to the flowcharts shown in FIGS. The operation of the outline vector detecting / reducing / smoothing / magnifying unit of FIG. 13 is shown in the flowchart of FIG.

Now, in step S11 of FIG. 21,
The CPU 71 reads the rough contour data given to the disk device 72 via the disk I / O 73 and reads it into the working memory area in the RAM 76.
Next, in step S12, the first smoothing and scaling processing is performed, and the data after the first smoothing processing is stored in the RAM 76. The details of the processing are omitted because they are the same as those in step S2 of FIG. 8 described in the related art. Thus,
After finishing the process of step S12, the CPU 72 performs the same-direction continuous vector detection / deletion process of step S13.

The contents of the process for detecting and deleting continuous vectors in the same direction in step S13 are shown in the flowchart of FIG. The continuous vector detection / deletion process is performed by inputting the data after the first smoothing. That is, the number of closed loops, the number of contour points for each closed loop, the coordinate value data string of the first smoothed contour points for each closed loop, and the additional information data string of the first smoothed contour points for each closed loop. Input and output contour point data after continuous vector detection / deletion processing.
The contour data after the continuous vector detection / deletion processing is shown in FIG.
As shown in FIG. 1, the number of closed loops, the contour point number table for each closed loop, and the coordinate value data string of the contour points after the continuous vector detection / deletion processing for each closed loop are configured.

Similar to the first smoothing, the continuous vector detection / deletion process is performed for each contour loop, and the process is advanced for each contour point in each contour loop.

22 will be described below. Since the flowchart of FIG. 22 describes the processing contents of one closed loop, the closed loop included in the input data after the first smoothing is input. 22 is repeated as many times as the number of times.

In step S101, the CPU 71 executes R
From the AM 76, the number of contour points of the contour loop to be processed is read into the working area 2080 in the RAM 76 shown in FIG. In step S102, it is determined whether or not the number of contour points in the contour loop is larger than 4, and if it is larger, the process proceeds to step S104, and if not, step S104.
Go to 103. The fact that the contour point is 4 here means that
It means that it is a rectangle, and that there is no vector with the same direction. Therefore, step S10
In 3, it is assumed that there is no contour point to be deleted on the contour vector loop currently being processed, and for the target loop, all the contour point data are output as they are, and the processing for that contour loop ends.

On the other hand, when the process proceeds to step S104,
Initialization necessary for proceeding the processing of a series of contour points on the contour loop is performed. The contents of this process are shown in FIG.
3 in detail.

In the initialization routine shown in FIG. 23, the area 2060 (RAM 76 in FIG. 27) for holding the equal inclination number indicating the situation where the contour vectors having the same inclination are continuous.
It is secured inside, and is referred to as an isometric number buffer hereinafter)
Is set to 5 as an initial value (step S201).

The meaning of the equi-tilt number will be apparent from the description below, but if it is simply explained, it is for counting the number of vectors having the same gradient.

In step S202, the contour point data three points before the start point in the loop is fetched as the previous point data. That is, the loop of interest is the i-th loop in the image,
When the total number of contour points in the loop is Li, the data of the _i- th loop L _i-2 vertex (see FIG. 11) is x _i Li _-2 , y _i L _i-2 and its corner points. Information (information indicating whether it is a corner point or not) is copied to the start point data storage area of the area 2000 in FIG. In step S203, similarly, the data x _iLi-1 and y _{iLi-1 of} the L _{i -1} vertex and its corner point information are copied to the current point data storage area 2000. With this, the last 2 points to be referred to when performing processing for the first contour point
Preparation for reading out one piece of contour data will be performed.

In step S204, the coordinate values of the vertices in the previous point data storage area 2010 (for convenience, x _k-1 , y _k-1
Will be written as. Therefore, here, x _k-1 = x _i
L _i-1 , and y _k-1 = y _i L _i-1 ) and the coordinate values of the vertices in the starting point data storage area 2000 (for convenience,
It will be expressed as x _k , y _k . Where x _k = x iLi
In and, y _k = and has a Yili) from _{a, △ x k = x k -x} k-1 △ y k = y k -y k-1 ... current side slope data by calculating the equation represented by △ x _k, it generates a △ y _k, to retain the current edge data storage area 2040 in FIG. 27.

In step S205, the front side data is updated by setting the current side data as the front side data.
The processing content will be described in detail in the flowchart of FIG.

In FIG. 24, in step S301,
Similarly, the contents of the area where the coordinate values of the contour point two points before and the corner point data on the contour loop with respect to the current point 2020 secured in the RAM 76 of FIG. The data is copied to the area holding the data of the contour points three points before in the area 2030. In step S302,
The content of the storage area 2010 of the previous point data is copied to the data area 2020 of the contour point two points before. Step S30
In 3, the contents of the current point data storage area 2000 are copied to the previous point data storage area 2010. In step S304, the contents of the isometric tilt number buffer 2060 are copied to the immediately preceding isometric tilt number buffer secured in the temporary storage area on the RAM 76 shown at 2070 in FIG. Step S30
In 5, the contents held in the current side data storage area 2040 are copied to the front side data storage area on the RAM 76 shown at 2050 in FIG. 27, and the series of processing for updating the front side data is completed. , Return to the parent routine. That is, the contour data is shifted one by one.

Thus, when the front side data is updated in step S205, the process proceeds to step S206. Step S2
At 06, as in steps S202 and S203, the data x _i L _i , y _i L _{i of} the L _i vertex and its corner point information are read and copied to the current point data area indicated by 2000 in FIG. The isotropic number is updated in step S207. The details of this processing are described in detail in the flowchart shown in FIG.

In FIG. 25, in step S401,
Just as in step S204, the current side inclination data is calculated from the current point data and the previous point data according to the formula, and
Current side data storage area 204 on RAM 76 shown in FIG.
Store in 0. In step S402, the region 2040 and 2050 of the RAM 76, whether the current edge data and the previous side data held respectively equal following the equation described _{above, △ x k-1 × m} = △ x k △ y k _-1 × m = Δy _k (where m is a real number) ... Is determined. Then, if the expression is satisfied, it is determined that the inclination of the current side is equal to the inclination of the front side, and the process proceeds to step S404. If not, the inclination is determined not to be equal and the process proceeds to step S403.

In step S403, the isotropic number is returned to 5, and the flow advances to step S405. In step S404,
The isotropic number is rewritten to the value obtained by subtracting 1 from the value at that time. In step S405, it is determined whether or not the current point is a corner point by referring to the area 2001 storing the corner point information of the current point in FIG. 27, and in the case of a corner point, step S406.
If it is a non-corner point, the process directly returns to the parent routine. In step S406, the isotropic number is rewritten to a value obtained by subtracting 1 from the value at that time, and the process returns to the parent routine.

In the routine of FIG. 25 described above, if the front side and the current side have the same inclination, the isometric tilt number is decremented by 1, and if the front side and the current side have different inclinations, the isometric tilt number is reset to 5. Is. Moreover, when the current point is a corner point, the number of points is further reduced by one.

Thus, step S20 in FIG.
If the isotropic number is updated in step 7, the process proceeds to step S208. In step S208, as in step S205, the routine shown in FIG. 24 is called to update the front side data by setting the current side data at that time as the front side data. Next, proceeding to step S209, the value of the unprocessed contour point counter 2080 in FIG. 27 is initialized with a value obtained by adding 3 to the number of contour points in the contour loop. This is because, in the present embodiment, three neighboring contour contour points before and after the contour point of interest are referred to. Next, in step S210, the value of the output contour point counter 2090 in FIG. 27 is initialized to 0, the series of initialization processing described above is completed, and the process returns to the parent routine.

Thus, step S10 in FIG.
The process of 4 is completed.

Then, the process proceeds to step S105.
In step S105, the start point on the contour loop (first
27), the coordinate values xi, yi and their corner point information are copied to the current point data area 2000 in FIG. In step S106, FIG.
The isosceles number updating routine described in 5 is called to determine whether the slopes of the front side and the current side are equal, and if they are equal, the isosceles numbers are respectively determined according to whether the current point is a corner point or not. 2 or 1
If they are not equal, the isotropic number is reset to 4 or 5 depending on whether the current point is a corner point or not. Step S10
In step 7, the front side data update routine described with reference to FIG. 24 is called to update the front side data by setting the current side data at that time as the front side data. Next, in step S108, the processed contour points are sequentially output. The details of this processing will be described in detail in the flowchart of FIG.

In FIG. 26, in step S501,
It is determined whether the isosceles number stored in the isosceles number buffer 2060 in FIG. 27 is positive, and if it is positive, step S5.
02, and if not positive, the process proceeds to step S506. In step S502, it is determined whether or not the front point is the corner point by referring to the corner point information 2011 of the front point in FIG.
If the previous point is a corner point, the process proceeds to step S503, and if it is a non-corner point, the process proceeds to step S504. Step S503
Then, it is determined whether or not the immediately preceding isometric number stored in the immediately preceding isometric number buffer 2070 in FIG. 27 is positive, and if positive, the process proceeds to step S504, and if not positive, the step S50.
Go to 6. In step S504, the contour point data 2030 three points before in FIG. 27 is output as processed contour point data. In step S505, the output contour point numerical value held in the output contour point counter 2090 in FIG. 27 is incremented by 1. In step S506, FIG.
2080, the unprocessed contour point number held in the unprocessed contour point counter is decremented by 1, and the process returns to the parent routine.

In the routine of FIG. 26 described above, when the isotropic number value at that time is 0 or negative, even if the contour point is deleted, the contour point is obtained after smoothing as compared with the case where it is not deleted. This means that the shapes are the same. Therefore, the contour point data three points before the target point (current point) is not output, that is, the contour point data is deleted.

On the other hand, when the isotropic number value at that time is positive, the determination is not made, and the contour point data three points before the point of interest (current point) is detected / deleted in the same direction. It is output as one of the processed contour points. However, even in this case, if the previous point of the attention point is the corner point and the isotonic number is 0 or negative at the time when the previous point is the attention point (that is, one point before). In addition, it is determined that the point 2 points before the corner point, that is, the point 3 points before the current point, is still the erasable contour point, and is not output.

Thus, when the process of step S108 is completed, the process proceeds to step S109. In step S109, the value of the unprocessed contour point number 2080 in FIG.
If it is larger (that is, if the point of interest is not the final point on the contour loop), step S
If not, the process proceeds to step S111. In step S110, the point of interest is moved to the next point on the contour loop, this point is set as the current point, the coordinate value and the corner point information are copied to the current point data area 2000 in FIG. 27, and the process returns to step S106. . Step S11
In 1, it is determined whether or not the value of the number of unprocessed contour points is 3 and 3
If (that is, the target point is just the final point on the contour loop), the process returns to step S105, and if not 3, the process proceeds to step S112. In step S112,
It is determined whether or not the value of the unprocessed contour points is positive. If the value is positive, it is determined that the point three points before the target point is not the final point on the contour loop, and the process proceeds to step S110. If not positive (that is, 0), the process proceeds to step S113. In step S113, all the contour points on the contour loop have been examined, and the value of the area 2090 in FIG. 27 is output as the number of contour points output as the contour points on the processed contour loop. A series of processing for one contour loop is completed.

In this way, the processing for detecting and deleting continuous vectors in the same direction in step S13 is completed. The processing result is
It is held on the RAM 76. In this process, the number of loops and the number of contour points do not increase, so if the data from the start point of each loop to the three points is temporarily saved in the temporary area, the data area of the first smoothing itself Can be used as the data area after this processing.

Next, the CPU 72 carries out the second smoothing in step S14 and the smoothed data output in step S15. The contents of these processes are as described in the prior art, so the details here will be described. I omit the description.

Thus, the outline vector detecting / reducing / smoothing / magnifying unit 13 shown in FIG. 1 is realized.

Incidentally, step S201 and step S403
Then, the isosceles number is reset to 5 when, after the slopes of the front side and the current side are changed and then the sides having the same slopes continue over the five sides, of the seven points forming the six sides. This is because the central point can be reduced. Further, when there is an angular point, the isocline number is additionally reduced by 1 due to the above-mentioned consideration that the number of consecutive sides having the same inclination may be reduced.

<Second Embodiment> In the above-mentioned embodiment, the contour points to be removed are
If not removed, it is removed after satisfying the same condition as the contour shape obtained by the second smoothing.

However, when the enlargement magnification is not so large, there may be a case where it is practically acceptable even if a slight difference in contour shape is recognized. In this case, in step S201 and step S403 of the above-described embodiment, a small change in contour shape is recognized by setting the value for resetting the isotropic number to a small value such as 4 or 3 instead of 5. In
More contour points can be deleted. If the isotropic number is set to 4, the slopes of the front side and the current side will change and then 4
When four or more sides with the same slope continue over the sides,
Of the six points forming the five sides, the third point from the start point side can be reduced. Further, in the case of setting the equi-tilt number to 3, at the same time, the points to be output in step S504 are set to the contour point data two points before, so that the slopes of the front side and the current side are changed, and then the three sides are processed. When three or more sides having the same slope continue, the four sides are composed 5
The central point of the points can be reduced.

Similarly, it is possible to further increase the number of reduction points by setting the isotropic number to 2 or 1 and setting the output point as the contour data of one point before. It is needless to say that the steps related to the contour point data and the like three points before which need not be referred to in these cases may be omitted.

<Third Embodiment> In addition to the above-mentioned embodiment, 2
For contour points between two different corner points, if the slopes of the side vectors existing between these corner points are all the same, no matter how many side vectors have the same slope continuously, the corner points between these corner points It is possible to delete a certain contour point without changing the contour shape even if the second smoothing is performed later.

For each contour loop, it is determined whether or not the side vectors between the corner points all have the same inclination, and if applicable, a process of deleting all the contour points existing between the corner points is added. May be. In this case, it is possible to reduce the number of contour points more than in the above-described embodiment without changing the contour shape.

<Fourth Embodiment> In the above, redundant contour points are deleted between the first smoothing and the second smoothing, but the present invention is not limited to this.

That is, after the second smoothing, the process of detecting / deleting continuous vectors in the same direction may be performed.

In this case, it does not matter whether it is a corner point or how many sides have the same slope. If two sides have the same slope, the middle point can be deleted. Is. The outline vector detection / reduction / smoothing / scaling means can be realized by simple processing, but the number of contour points to be subjected to the second smoothing processing does not decrease. no change.

<Fifth Embodiment> The outline vector detecting / reducing / smoothing / scaling means can be realized before the first smoothing or in the first smoothing.

For example, in the first smoothing method already proposed by the applicant of the present application, as described above, the contour edges (horizontal vector or vertical vector) vector of each rough contour data are sequentially focused. For each target contour edge vector, up to three vectors before and after that (ie,
The pattern is divided by a combination of lengths and directions of mutually continuous side vectors of three in front of the side of interest, the side of interest itself, and a total of seven side vectors of three after the side of interest).
For each case, the contour points after the first smoothing, which are the results of the first smoothing on the target side, are defined.

Among the combinations of the lengths and the directions of the seven continuous contour side vectors, FIG. 28, FIG. 29 and FIG.
As indicated by 0, it is clear that the contour points after the first smoothing are the seven consecutive contour points described in the above embodiment, and the six side vectors formed by these all have the same slope. There is something like this. That is, in FIG. 28, the length of the target side vector is 1 or 2, all three front and rear sides have the same length as the target side vector, and the target side vector, and two sides before and two sides behind. The direction of the side vector is the same, and the directions of the front and rear sides are the same as those of the front side, the front side, and the rear side. In FIG. 29, the length of the target side vector is 1, the lengths and directions of the side vectors before and after the two sides are the same as the side of interest, and the front and rear sides and the side vectors before and after the three sides,
This is the case when the length and direction are the same. In FIG. 30, the length and direction of the target side vector are the same as the side vectors before and after the two sides, and the lengths of all the side vectors before and after, the three sides before, and the sides after three are 1 and the directions are the same. This is the case. When these patterns are detected, the contour points for the first smoothing are not defined, that is, the contour points after the first smoothing for the target side vector are assumed to be deleted at this point and the process proceeds. It may be configured as follows.

Also in this case, if the slight difference in contour shape caused by the subsequent processing is allowed between the case of deleting and the case of not deleting, for example, in the patterns shown in FIG. 31 and FIG. At this point, the removal (not defined) of the contour points after the first smoothing for is effective. In FIG. 31, the lengths of the seven vectors are all 1, and
This is a case in which the direction of the target side is the same as the direction of the side vectors before and after the two sides, and the directions of the front side and the side after the three sides are the same, and the directions of the rear side and the side before the three sides are the same. . In FIG. 32, the lengths of the seven vectors are all 1, the directions of the side vectors before and after the two sides are both opposite to the side vector of interest, and the three sides before, the front side, the rear side, and the three sides are three. This is the case where the directions of the edge vectors after the edge are all the same.

It should be noted that it is possible to obtain more reduction patterns by referring to more sides without limiting the number of neighboring sides to be referred to to three at the front and back.

<Sixth Embodiment> In the above description of the embodiments, the input of the outline vector detection / reduction / smoothing / magnifying means is the binary image acquiring means and the outline extracting means in the preceding stage. However, the present invention is not limited to this, and for example, a similar contour vector extracted outside this apparatus may be obtained via a known interface realized by an I / O port or the like, or in advance. Alternatively, the contour vector data stored in the disk device by other means may be input at a later date by another instruction via the disk I / O or the like.

Also, outline vector detection / reduction /
The output of the smoothing / magnifying means is explained as the binary image reproducing means, but the output is not limited to this, and it is output in the form of contour data outside the device via the interface or is stored in the disk device. Even if it is an output circuit that operates, it does not matter.

<Seventh Embodiment> FIGS. 33 to 35 are block diagrams showing the case where the present invention is applied to a facsimile apparatus.

FIG. 33 is a block diagram in which the present embodiment is applied to a facsimile machine on the receiving side. The code transmitted by the MH code or the like is decoded to create input binary image data, and outline extraction and smoothing processing are performed. I do. The binary image regenerated by this processing is output to recording paper or the like by the recording device or displayed by a display device (not shown).

FIG. 34 is a configuration diagram in which this embodiment is applied to a facsimile machine on the transmission side. An image input by a scanner or the like is binarized to generate input image data, and the outline extraction and smoothing processing of the above-described embodiment is performed. The binary image regenerated by this processing is temporarily stored in the image memory, encoded by the encoder into the MH code, and transmitted.

FIG. 35 is a block diagram when this embodiment is applied to both input images of transmission and reception. Although a combination of the above two examples, the selector is controlled by the transmission / reception control circuit, and the input / output of the outline processing (outline extraction and smoothing processing) is determined by transmission / reception. Here, in particular, it is possible to select a reading unit as the binary image acquisition means and configure it as a recording apparatus as the binary image output means. ) Is possible.

As described above, continuous short vectors having the same slope are detected between the outline extracting means (step) and the binary image reproducing means (step) in the conventional example, and these short vectors are connected. By providing means (steps) and reducing the number of vectors to be processed after the means (steps), the processing time required for the entire processing can be reduced.

Further, by reducing the memory capacity required for processing, there is an effect of reducing the device cost.

The present invention may be applied to a system composed of a plurality of devices or an apparatus composed of a single device. Further, it goes without saying that the present invention can be applied to the case where it is achieved by supplying a program to a system or an apparatus.

[0109]

As described above, according to the present invention, when the contour points of the obtained binary image are passed to the subsequent process, the load of the subsequent process can be reduced as much as possible.

[0110]

[Brief description of drawings]

FIG. 1 is a basic configuration diagram of a binary image processing apparatus according to an embodiment.

FIG. 2 is a diagram for explaining reading of a binary image by a raster operation type.

FIG. 3 is a diagram showing a principle of extracting a contour vector.

FIG. 4 is a diagram illustrating a manner of extracting an outline from a raster scanning binary image.

FIG. 5 is a form diagram of outline data output from outline extracting means.

FIG. 6 is a functional block diagram of an outline smoothing / magnifying unit.

FIG. 7 is a block diagram of outline smoothing / scaling means and outline vector detection / reduction / smoothing / scaling means.

FIG. 8 is a flowchart showing an outline of outline smoothing / magnifying processing.

FIG. 9 is a diagram illustrating a processing operation of first smoothing.

FIG. 10 is a diagram showing a configuration of a conventional binary image processing device.

FIG. 11 is a morphological view of contour data after the first smoothing.

FIG. 12 is a diagram for explaining the content of second smoothing processing.

FIG. 13: Outline vector detection in the embodiment
It is a functional block diagram of a reduction / smoothing / scaling means.

FIG. 14: Detection of continuous vectors in the same direction in the embodiment
It is a figure explaining the operation principle of a deletion means.

FIG. 15: Detection of continuous vectors in the same direction in the embodiment
It is a figure explaining the operation principle of a deletion means.

FIG. 16: Detection of continuous vectors in the same direction in the embodiment
It is a figure explaining the operation principle of a deletion means.

FIG. 17: Detection of continuous vectors in the same direction in the embodiment
It is a figure explaining the operation principle of a deletion means.

FIG. 18: Detection of continuous vectors in the same direction in the embodiment
It is a figure explaining the operation principle of a deletion means.

FIG. 19 is a detection of continuous vectors in the same direction in the embodiment.
It is a figure explaining the operation principle of a deletion means.

FIG. 20: Detection of continuous vectors in the same direction in the embodiment
It is a figure explaining the operation principle of a deletion means.

FIG. 21 is a diagram illustrating outline vector detection in the embodiment.
It is a flow chart which shows an outline of reduction, smoothing, and magnification processing.

FIG. 22 is a flowchart showing the flow of processing for detecting and deleting continuous vectors in the same direction for one contour loop in the embodiment.

FIG. 23 is a flowchart showing in detail the flow of processing for detecting and deleting continuous vectors of the same direction for one contour loop.

FIG. 24 is a flowchart showing in detail the flow of processing for detecting and deleting continuous vectors of the same direction for one contour loop.

FIG. 25 is a flowchart showing in detail the flow of processing for detecting and deleting continuous vectors of the same direction for one contour loop.

FIG. 26 is a flowchart showing in detail the flow of processing for detecting and deleting continuous vectors of the same direction for one contour loop.

FIG. 27 is a diagram showing the contents of a work area required for the process of detecting and deleting continuous vectors in the same direction.

[Fig. 28] Fig. 28 is a diagram illustrating another example of a deletion target of a contour vector.

[Fig. 29] Fig. 29 is a diagram illustrating another example of a deletion target of a contour vector.

[Fig. 30] Fig. 30 is a diagram illustrating another example of a deletion target of a contour vector.

[Fig. 31] Fig. 31 is a diagram illustrating another example of a deletion target of a contour vector.

[Fig. 32] Fig. 32 is a diagram illustrating another example of a deletion target of a contour vector.

FIG. 33 is a configuration diagram of an example in which the present invention is applied to a receiving facsimile.

FIG. 34 is a configuration diagram of an example in which the present invention is applied to a transmission facsimile.

FIG. 35 is a configuration diagram of an example in which the present invention is applied to a transmitting / receiving facsimile apparatus.

[Explanation of symbols]

11 Binary image acquisition means 12 Outline extraction means 13 Outline vector detection / reduction / smoothing / magnification means 14 Binary image reproduction means 15 Binary image output means 21 Magnification setting means 24 Consecutive vector detection / deletion means 71 CPU 76 RAM 75 I / O 72 Disk device 74 ROM

Claims (6)

[Claims] 1. A contour vector acquisition means for acquiring contour vector data extracted from a binary image, and the vector group based on the inclinations of a predetermined number of vector groups adjacent to each other acquired by the contour vector acquisition means. An image processing apparatus comprising: a detection unit that detects an unnecessary vector from among the above, and a deletion unit that deletes the vector detected by the detection unit. 2. The contour vector acquiring means comprises a means for acquiring a binary image and a contour vector extracting means for extracting contour vector data from the binary image. The image processing device according to item 1. 3. The image processing apparatus according to claim 1, wherein the detection unit includes a determination unit that determines whether or not a predetermined number of vectors having the same direction are continuous. 4. An unnecessary vector is detected from the vector group based on the process of acquiring the contour vector data extracted from the binary image and the inclinations of the acquired predetermined number of adjacent vector groups. An image processing method, comprising: a step; and a deletion step of deleting the detected vector. 5. The process of acquiring the contour vector includes a process of acquiring a binary image and a process of extracting contour vector data from the binary image. The described image processing method. 6. The image processing method according to claim 4, wherein the detection step includes a step of detecting how many vectors having the same direction are continuous.