JPS6353587B2 - - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a digital image processing method, and particularly to a digital image processing method that is optimal for extracting a specific image from among digitized images.

In the past, when a specific image was extracted from a graphic display image and the extracted image was inserted into another image to create another combined image, this was done manually using paper and scissors. Based on the results of this work, using the keyboard and light pen,
Specific images were extracted from CRT images and replaced.

In recent years, it has been considered to automatically perform such tasks, but the key to this task is the process of automatically cutting out a specific figure from an image. However, this process has become technically difficult, and the following methods have been published so far.

(1) Input each point on the contour of the cut-out figure into the computer in detail (for example, by connecting the figure lines with a straight line using a light pen).

(2) Input the minimum necessary points on the contour of the cut-out figure into the computer (for example, specify and input three points on the figure with a light pen). In this case, the distance between input points is determined by image data analysis.

However, although the former method is easy to implement because computer processing is simple, it has the disadvantage that it is not possible to accurately cut out the figure unless a large number of points on the figure are input.

In addition, although the latter method requires fewer points to be input and therefore has excellent operability and is practical,
The disadvantage is that the method of image data analysis requires ingenuity.

SUMMARY OF THE INVENTION An object of the present invention is to provide a digital image processing method that can automatically create the outline of a figure while minimizing the number of input points for cutting out the figure.

The present invention creates a luminance level state diagram of the processing target based on known data of the contour of the imported figure, analyzes the amount of change, searches for the next relay point, and sequentially forms the contour. It is something.

FIG. 1 is a schematic configuration diagram showing an embodiment of the present invention.

A central processing unit 1 that executes a processing program to be described later is the core, an external storage device 2 is connected to the central processing unit 1 via an interface 3, and an image display device 4 and a keyboard 5 are connected via an interface 6. Ru. Furthermore, a light pen 7 is connected to the keyboard 5 via an interface 8.

The image data to be processed (digitized image data) is stored in advance in the external storage device 2 and can be referenced from the central processing unit 1 whenever necessary. The image data recorded in the storage device 2 is displayed on the screen of the image display device 4 via the central processing unit 1. Image display device 4
The light pen 7 and various function keys on the keyboard 5 are used to cut out a specific figure in the image on the screen. FIG. 2 shows an example of the arrangement of this function key.

The mode changeover switch 9 is for switching between automatic and manual mode. The function key 10 is for defining the starting point and defines the starting point of the light pen at any position on the screen, the function key 11 is for defining the intermediate point of the outline, and the function key 12 is for redefining the intermediate point, so that the displayed image after processing is This is used when the desired state is not achieved. Furthermore, the function key 13 is for end point definition and is used to define the end point of automatic contour processing.

First, place the light pen 7 at the desired starting point on the outline of the figure you want to cut out on the CRT and press the function key 10. Next, place the light pen 7 at another point on the same contour line, and define this coordinate point as a relay point by pressing the function key 12. The input of this relay point shall be adjusted by the operator so that it is within the data analyzable range. It is also assumed that the end point of contour processing is defined as the end point, the light pen 7 is placed at this point, and the function key 13 is pressed. A boundary line is drawn between the starting point and the relay point by analyzing the image data (if it is not indicated by a line, changes in color shading are recognized as a boundary line)
is automatically determined and displayed as an image. The operator is
The operator looks at this display result and determines whether it is appropriate as the outline of the desired figure. If it is determined to be appropriate, specify the next relay point. If it is determined that the shape is not valid, the relay points are redefined so that the outline of the figure is accurately displayed. It can be seen that by repeating the process of inputting relay points and determining the boundary of a figure in this way, it is possible to easily determine all the lines to be drawn for the figure. Moreover, the input of relay points is minimal, depending on the complexity of the figure, and manual operations are simplified.

FIG. 3 is a block diagram showing the internal configuration of the image processing device 4. As shown in FIG. Data to be processed from the central processing unit 1 is sent to the image display device 4 via the interface 6. Controller 1 of image display device 4
4 and interface 15 to image memory 16 . This image memory 1
When processing a color screen compared to a self-black screen, the capacity of 6 is required to be three times larger than that of a black screen because the color is composed of three primary colors. The image memory 16 is capable of storing gradation levels of 1 bit to 8 bits per pixel, and the contents of the image memory 16 are configured to be displayed as images on the display screen 41 as they are. The graphic memory 17 is for displaying straight lines and curves on the screen, and the data is
The data is transferred through the interface 6 and via the controller 10. The display screen 41 is configured to display the contents of the image memory 16 and the graphics memory 17 in an overlapping manner. With such a configuration, the image data to be processed can be stored in advance in the image memory 16, and the contents of straight lines and curves necessary for the work can be sequentially displayed in the graphic memory 17. In order to cut out a specific figure in the image on the screen 41 of the image display device 4, the coordinate values specified by pressing the light pen 7 and various function keys 10 or 11 on the keyboard 5 are centrally processed. It is taken into the device 1. Also, when redefining, the function key 12 is used. Function key 1
3 is used, the boundary line between the predefined relay point and the starting point is processed. Also, when the shape is complex and cannot be determined by calculation,
This is done by operating the mode changeover switch 9 to switch between the process of connecting two points with a straight line.

FIG. 4 is a flowchart showing the series of processes described above by the central processing unit 1. First, a starting point is defined using the function key 10, which is executed in detail according to the flowchart shown in FIG. 5, and the starting point coordinate values are imported into the central processing unit 1 side. Next, the relay point is defined by operating the function key 11 and executed according to the flowchart shown in FIG. 6, and the value of the relay point is read from the screen. Based on the above information import results, the central processing unit 1 performs processing to determine a curve between two points (starting point and relay point, relay point and next relay point), that is, between the two input coordinates. Let's move on to processing the boundaries. This process is the most characteristic feature of the present invention, and will be explained in detail based on the drawings.

As shown in FIG. 7, a boundary line is determined between the starting point (or the previous relay point) and the first relay point (or the subsequent relay point) by repeating the process n times. After the starting point i ₀ is defined, the first intermediate point i _o
The boundary line up to i follows the path of i ₁ , i ₂ , i _n+1 , i _o-1 , but first, i ₁ is determined with i ₀ as the base point, and then i ₁ is determined as the base point. i ₂ is determined as i _o
Routes are determined sequentially until reaching . This processing result is displayed in black and white mode on the screen of the image display device 41 (displayed over the original screen). The operator then looks at the display screen and judges whether it is good or bad. If the graphic data is not normal or the designation of the relay points is not valid, the processing will not be fixed and the relay points cannot be easily reached (e.g. when the relay points are sparsely arranged and the curve shape is complex, etc.) )Sometimes. In such a case, the central processing unit 1 determines that there is an error, and operates the function key 11 to redefine the definition. If the first relay point is i ₀ and the next relay point is i _o , the internal area of the figure is considered to be X, and the external area is considered to be in contact with several areas. Figure 8 shows a case where the external area consists of two areas, Y and Z (for example, when adjacent areas have different color tones, or there is a change in shading (change in brightness) in the color, forming a boundary) An example of this is shown. Considering the distribution of digital data on the bb section of FIG. 8, the boundary between the X and Y regions can be determined from the distribution of digital data corresponding to the density difference (x, y) between the two, as shown in FIG. Also, c-c in Fig. 8
The cross-sectional distribution can be determined from the distribution of digital data in which the boundary between the X and Z regions corresponds to the density difference (x, z) between the two regions, as shown in FIG. In this way, in a figure in which there is a clear difference in digital data distribution between regions, the boundary line can be easily determined. Note that when the above-mentioned redefinition is executed, the position where the light pen 7 is placed is determined so that the interval between the relay point and the next relay point becomes narrower. The boundary line processing will be described in detail below with reference to the drawings.

FIG. 11 is a flowchart when boundary line processing is performed by the central processing unit 1, FIG. 12 shows a boundary line model diagram, and FIG. 13 is a grayscale level diagram when searching for a boundary line.

In Figure 12, when points i _n-1 and i _n are already determined points, as known data
i _n-1 , i _n , r are taken in. Next, if the i _n+1 point is the desired point, then
Consider a region spanning 180 degrees, that is, a semicircle with a radius r centered at point i _n (when defining a relay point, there is no need to consider the direction of the entire circle since no retrograde figure is included). Search this semicircular area (j ₁
→j _o ′), points where a difference in shading occurs, such as y or z in Figure 13, are found,
It can be determined that the j _n ′ point is the boundary line. This j _n ′ point will be used as a relay point i _n+1 to form a boundary line. In this way, all boundaries are determined by sequentially growing the boundaries in predetermined i steps.

Next, a specific example of processing will be shown.

FIG. 14 shows a figure A surrounded by three areas B, C, and D, and describes the gray level of each area and the figure A when it is digitized.
Further, FIG. 15 shows a digitized version of the gray level shown in FIG. 14 for one screen. In this example, the pixels are large, so the figure A in FIG. 14 is slightly distorted, but in reality, the pixels are very fine, so the digital image can reproduce the natural shape. In FIG. 15, the horizontal axis is called a pixel, and its coordinate values are 1 to 22. Further, the vertical axis is called a line, and its coordinate values are 1 to 18. Each point shall be represented by coordinates (pixels, lines). The boundary is also displayed using two coordinates ((pixel 1, line 1), (pixel 2, line 2)). The procedure for cutting out figure A from a digital image will be described below.

(1) Digital image information shown in FIG. 15 (the gray level of each pixel is stored in the external storage device 2, and also stored in the image memory 11 in the image display device).

(2) The contents of the image memory 11 are displayed on the screen 41.

(3) Input one boundary on the screen using the light pen 7 and press the start point definition function key 10 to define the start point.

(4) Next, input the boundary of the relay point with a light pen and press the function key 11 to define the relay point.

(5) Next, the boundaries connecting the starting point and relay points are determined by computer processing. The boundaries are found sequentially as follows.

Starting point a {(7, 7), (7, 8)} t ₁ {(8, 7), (8, 8)} t ₂ {(9, 7), (9, 8)} t ₃ {(9 , 7), (10, 7)} t ₄ {(9, 6), (10, 6)} t ₅ {(9, 5), (10, 5)} t ₆ {(10, 4), ( 10, 5)} t ₇ {(11, 4), (11, 5)} t ₈ {(11, 5), (12, 5)} t ₉ {(11, 6), (12, 6)} t ₁₀ {(12, 7), (12, 6)} Relay point b {(13, 7), (12, 6)} To find the boundary point between these two points, use various logic. Prepare it and configure it so that you can use it while selecting the logic.
Although the configuration is omitted here, in the case of this example, the following logic was selected and used. From a, t ₁ to t ₆ , the same boundary conditions as point a (boundary between level 1 and level 5)
While selecting , perform operations to extend the link lines one by one. Next, at _t7 , it is about to branch into two boundaries: levels 1 and 3 and levels 3 and 5. Here, it is determined that the boundary between levels 3 and 5 is closer to the boundary of point a than the boundary between levels 1 and 3. Below, from _t7 to _t10 , the boundary between 3 and 5 is taken. Since we arrived at the relay point in this way, we can conclude that this is one of the correct answers. Such processing can be performed automatically by storing the logic in the computer in advance. The link lines from the starting point to the relay point are sequentially stored in the graphic memory 12. Therefore, linking lines are sequentially displayed on the screen 41. Thereafter, when the same operation is repeated and the final intermediate point is input, the end point is defined using the function key 13, and all link lines are closed. It can be seen that any figure can be easily cut out using this method.

In the above description, the case where there is one boundary line has been explained as an example, but the processing when there is a branch point will be explained. For example, in the case of the branch at point X shown in Figure 8,
The relay point may be defined as point X, and thereafter, the above-described boundary line processing may be executed for each branch line.

The present invention is suitable for application to digital image processing such as image cutting out for automating color plate making or cutting out image evaluation areas.

As is clear from the above, according to the present invention, a specific figure in an image can be easily cut out with a small number of input points.

[Brief explanation of drawings]

1 is a schematic configuration diagram showing an embodiment of the present invention, FIG. 2 is a key arrangement diagram of the keyboard 5 shown in FIG. 1, and FIG. 3 is a memory configuration diagram of the image display device shown in FIG. FIG. 4 is a processing flowchart of the present invention, FIG. 5 is a processing flowchart of starting point definition in the present invention, FIG. 6 is a processing flowchart of relay point definition in the present invention, and FIG. 7 is an explanatory diagram of boundary line processing. FIG. 8 is an explanatory diagram of boundary line processing when there is a brightness change in the figure; FIGS. 9 and 10 are brightness change diagrams in the bb section and the cc cross section of the figure in FIG. 8;
FIG. 11 is a flowchart of the boundary line determination process between relay points in the present invention, FIG. 12 is an explanatory diagram of the boundary line determination process between relay points in the present invention, and FIG. 13 is a change in brightness depending on the direction of the figure in FIG. 12. 14 is a figure with a digital level showing an embodiment of the present invention, and FIG. 15 is a data layout diagram showing the figure of FIG. 14 at a digital level. 1...Central processing unit, 2...External storage device,
3, 6, 8, 15, 16...interface,
4... Image display device, 5... Keyboard, 7...
Light pen, 9...mode changeover switch, 10,
11, 12, 13...Function key, 14
...Controller, 16...Image memory, 1
7...Graphic memory.

Claims (1)

[Claims] 1. In an image processing method that cuts out a specific figure from a digital image and displays it on the screen of a display device, two arbitrary points on the boundary line that constitutes the outline of the figure are specified on the screen and input. A digital image processing method characterized in that a boundary line connecting the two points is defined by a change point of the brightness level state of the figure, and a continuous boundary line between the two points is displayed on the display device. . 2. In an image processing method that cuts out a specific figure from a digital image and displays it on the screen of a display device, two arbitrary points on the boundary line that constitutes the outline of the figure are specified and input on the screen, and the A boundary line connecting the two points is defined by the change point of the brightness level state of the figure, and a continuous boundary line between the two points is displayed on the display device, and when the display screen is judged to be inappropriate, the definition A digital image processing method characterized by redefining the advancing side point of the two points and redetermining the boundary line to be displayed.