JPH0877336A - Area extracting method - Google Patents

Abstract

PURPOSE: To reduce the input burden on an operator in the case of image cutting and to exactly extract an object area at high speed concerning the method for extracting any desired area out of a color image or continuous gradation images. CONSTITUTION: A source image stored in a storage means is displayed on a display means, an outer frame area is instructed and inputted in the shape of a band by a pointing means so as to include a contour part under displaying (S10 and S11), the difference from an adjacent picture element in a color space is calculated concerning the picture element of the source image within the outer frame area, and a partial image connecting picture elements showing the maximum difference and defining it as the contour line is defined as the extracting area in the source image (S12-S17).

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method for extracting a desired area from a color image or a continuous tone image, and in particular,
The present invention relates to a region extraction method suitable for reducing the burden of operator instructions and reducing the calculation time in the region extraction processing.

[0002]

2. Description of the Related Art Conventionally, a digital image processor controlled by a computer partially combines a plurality of still images to create a composite image, or selects only a necessary portion on a screen to obtain a desired position. In the process of moving to and editing, there is, for example, the following method as a method of extracting the image area to be processed. A method in which the operator traces and specifies the contour of the target area using a coordinate input device such as a tablet, and extracts the area within the contour. In this case, the target area is partially enlarged so that the contour can be accurately traced with the pen of the tablet so that the contour can be accurately input. A method in which an operator specifies the vicinity of a portion to be cut using a pointing device, performs a two-dimensional calculation process on a planar area in the specified vicinity to detect a contour, and extracts a region within the detected contour.

As the latter method, for example, an operator specifies a rough contour line while roughly tracing the contour line of a target area by using a pointing device, obtains edge strengths of original image pixels in the rough contour line, and then calculates the contour. There are known methods for detecting. As a specific example, referring to the edge strength obtained for the original image pixels in the rough contour line, the center line extraction processing is applied until the number of pixels to be deleted disappears with the edge strength equal to or less than the preset threshold as the deletion target. While increasing the threshold value, the centerline extraction processing for the rough contour line is repeatedly applied until the thinning result of the line width of 1 pixel is finally obtained, and the obtained thinning result is used as the contour line. There is a method in which an image is used as an extraction region in an original image (Japanese Patent Laid-Open No. 233810/1993).

[0004]

However, in the above method, since the operator directly specifies the cutting line, the operational burden is large and it takes time.
It also has the drawback of lacking accuracy. On the other hand, in the above method, the two-dimensional images of interest move one after another, and each of them requires an operation process such as differentiation and binarization.
It has a drawback that it is inferior in processing speed.

The present invention has been made under the circumstances as described above, and an object of the present invention is to reduce the input load on the operator when cutting out an image and to extract a target area at high speed and accurately. An object is to provide a region extraction method.

[0006]

According to the present invention, storage means for storing image data of a color image, display means for displaying the color image, and indication of arbitrary coordinates in the displayed color image, The present invention relates to an area extraction method for an image processing apparatus having input pointing means, and the above object of the present invention is to display the original image stored in the storage means on the display means, and to include a contour portion being displayed. As described above, the pointing means is used to input and input the outer frame area in a band shape, the difference between the pixel of the original image in the outer frame area and the adjacent pixel in the color space is obtained, and the pixel showing the maximum difference is connected to form the contour. This is achieved by using a partial image as a line as an extraction region in the original image.

Further, storage means for storing image data of the continuous tone image, display means for displaying the continuous tone image, and designation and input of arbitrary coordinates in the displayed continuous tone image. In an area extraction method of an image processing apparatus including pointing means, the original image stored in the storage means is displayed on the display means, and the outer frame area is displayed by the pointing means so as to include the contour portion being displayed. Is input in the form of a strip, and the difference in density between the pixels of the original image in the outer frame area with adjacent pixels is obtained, and the partial image in which the pixels showing the maximum difference are connected to form a contour line is the extraction area in the original image. Is achieved by

[0008]

According to the present invention, the contour line of the extraction area is automatically recognized only by pointing and inputting the outer frame area in a strip shape by the pointing means so as to include the contour portion being displayed. Further, in the instruction at that time, the band-shaped line is automatically displayed by indicating the main points, and therefore the operator does not need to trace the outline. Further, the contour line is detected by calculating the difference between the feature value of the pixel of the original image in the outer frame area and that of the adjacent pixel, and connecting the pixel showing the maximum difference to form the contour line. The contour can be detected with. At this time, the outer frame area is scanned in a cross pattern to calculate the difference, thereby further increasing the speed. Further, when the pixel showing the maximum difference cannot be detected, the center line of the outer frame area is used as the contour line, so that it is necessary to perform arithmetic processing such as a two-dimensional filter even on a portion where the difference from the background is unclear. Disappeared
The contour intended by the operator can be recognized by a simple calculation.

[0009]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS An image processing system which is a premise of the present invention will be described with reference to the drawings. FIG. 1 shows an example of the overall configuration of an image processing system. A plurality of editing workstations 10 are connected to each other via Ethernet, and one workstation 10 among them is relatively connected. A galley printer 20 for performing low-quality galley printing is connected. A workstation 30 for a server having a data server and a recorder server function is further connected to the Ethernet, and a mount input device 40 for reading and inputting a layout mount for printing is attached to the server workstation 30. A plurality of color scanners 50, 50 that are connected via the input controller 40A and that read and input a color image or a monochrome image of patterns, characters, figures, etc. by color separation are input via the plurality of input controllers 50A, 50A. It is connected. Further, the server workstation 30 has a high-quality image through an output synchronization buffer 70 which has functions of decompression, half-shading, merging (line drawing and continuous tone (monotone) image), and buffering for synchronization. Printer 6 for outputting
0 is connected.

FIG. 1 shows an example in which a plurality of editing workstations 10 and one server workstation 30 are systematically combined. As shown in FIG. 2, one editing / server workstation 30A is used. It is also possible to have a stand-alone configuration with. In addition, other information (for example, LAN (Lo
(Cal Area Network) information, information from other computer systems, etc.) and processes it. Although the output synchronization buffer 70 is inserted between the server workstation 30 and the film printer 60 in the configuration examples of FIGS. 1 and 2, it may be incorporated in the film printer 60. .

The editing workstation 10 and the server workstation 30 can take various forms depending on the system configuration, but here, for convenience, an example of the same hardware configuration is shown in FIG. 3 in detail. The workstation 30 (or 10) is a CPU 301 that controls the entire system.
And a hard disk 302 for storing necessary information, a CRT 303 as a display means, a keyboard 304 and a mouse 305 as an input operation means, a digitizer 306, a pointing means such as a trackball and a joystick, and a memory. A floppy disk (FD) 307 as a means can be loaded.

FIG. 4 is a block diagram showing the overall configuration of the system shown in FIG. 1. The mount information KS (1 bit) read by the mount input device 40 is sent to the workstation 30 via the input controller 40A. Has become
C of four colors read by a plurality of color scanners 50, 50
MYK (Cyan, Magenta, Yellow
w, Black) color information CL1, CL2 (= 3)
2 bits; or K monocolor information = 8 bits) is sent to the workstation 30 via the input controllers 50A, 50A. Input controller 40A (50A)
Is designed to simultaneously perform high-density data processing for high image quality processing and coarse-density data processing for display, etc. in parallel at the same time, achieving high speed as a whole and efficient data processing. Has been realized. Input controller 40A,
50A has the same configuration, and the thinning unit 401 repeatedly performs thinning in a feedback manner, for example, 1/2, 1 /
Input data K in integer ratio such as 3, ..., 1/6, 1 / n
Although S (CL1, CL2) is thinned out, naturally, thinning out is not executed for the high density data required for high image quality output. In the input controller 40A (50A), data compression is performed by the compression unit 402 together with data thinning,
The thinned and compressed data is temporarily stored in a buffer (not shown).

Data (1 bit (line drawing information), 8 bits (mono color), 32 bits (full color)) temporarily stored in the input controller 40A (50A) is input to the server workstation 30 as input information INS. External information EXS from another personal computer or the like connected to the external system is also input to the server workstation 30. The server workstation 30 includes a scan server 320 for converting the format of each input information and registering an image, and further includes a recorder server 310 for managing an output job. Allocation information from the recorder server 310 is provided. (Characters, figures, tables, etc.
Information for specifying the layout and size of photographs, etc.) PSD
Is input to the raster image processing unit (PSRIP) 312. Further, the server workstation 30 is provided with a data disk 311 for storing image data, and the image data IGS read from the data disk 311 is stored in the image replacement unit (Open PrePress Int).
erface) 313. Image replacement unit 3
The image data IGSA replaced in 13 is input to the raster image processing unit 312, and the raster image processing unit 3
The image is converted into a raster image at 12, and the pattern or the like in the image data is converted into a halftone dot. If necessary, the data-compressed raster data RD is input to the output synchronization buffer 70. The output synchronization buffer 70 synchronizes the data output with the printing speed of the film printer 60, performs necessary expansion on the data-compressed data, and further performs merging and halftone dot transmission to the film printer 60. To do.

FIG. 5 shows the detailed software configuration of the server workstation 30, and the input information INS.
The external information EXS is input to the format conversion unit 321 in the scan server 320, the format-converted data is used to create an image for display (display image) by the display image creation unit 322, and an image for icon (icon image). ) Is created, and the image registration unit 324 performs image registration processing. Registration data SS of the display image and the icon image output from the scan server 320
Although D is input to the database manager 330 and data is stored therein, the image data IG is stored in the data disk 311.
Is input to the allocation data (PostScript D
ata) PS is input to the recorder server 310 that manages output jobs. Database manager 330
Includes a large attachment module 340 for imposing each page, an image processing module 350 for performing line drawing preprocessing (noise removal, rotation, etc.), retouching and clipping processing of a continuous tone image, a mount layout, A plate collection module 36 for collecting image data such as image layout, figure generation, and shading
0, a data management module 370 that performs output job management and data management, and an operation data control module 380 that performs color editing, hatching editing, ground stop registration, and the like are connected by software.

Here, the processing operation of the database manager 330 will be described with reference to the flowchart of FIG. 6. When the registration data SSD such as the mount and the part image transmitted from the scan server 320 is input (step S1). Image data IG is stored in the data disk 311 via the database manager 330 (step S2). At the time of data correction, the database manager 330 reads out the image data from the data disk 311, and the image processing module 350 performs retouching (correction work for the purpose of correcting color, gradation, scratches, etc.), dust removal, cutout, etc. Manager 330
The data is stored in the data disk 311 via (step S3). Next, the database manager 330 performs plate-collection processing (arranges a mount, arranges each part along the mount, and performs character processing to synthesize parts such as characters, figures, and images according to the designated layout). However, the database manager 330 first reads the image data IG from the data disk 311, performs the plate collection operation by the plate collection module 360, and then stores the processed data in the data disk 311 via the database manager 330. Store. And further data disk 3
Image data IG is read out from 11, and large paste module 3
40, the imposition instruction and work are performed, the imposition data is stored in the data disk 311 via the database manager 330, and the imposition is performed (step S
5) Next, the data is extracted from the data disk 311 to the database manager 330, and the recorder server 310
The output process is performed in step S6.

FIG. 7 shows the target of input data (binary data,
The relation between layout information, bitmap data, continuous tone image) and process (input, plate collection / editing, output) is shown as data and flow of functions that the operator is aware of. The data is read by the mount input device 40 or the color scanner 50 to remove dust, retouch, rotate,
Enlargement / reduction processing is performed to obtain image editing data, and P
The ostScript information is RIP (Raster Im
image processor; data expressed in a page description language such as PostScript is developed and converted into bitmap data or the like) to be image editing data. Also, the bitmap data is format-converted by a filter to become image editing data, and the continuous tone image is read by the color scanner 50 or directly input to be retouched, cut out, and subjected to image processing to become image editing data. . The keyword, image name, and job name as the data management information are also incorporated in the image editing data. Image editing data is processed such as mount processing (closed area automatic recognition), drawing, object editing, composition, transformation, rotation, shading / attribute change, layout, photo embedding, etc., history image display change, distributed editing , Data saving is executed. After each of the above processes, color separation, spot color plate, and trapping are performed, and then imposition is performed page by page, and plate collection / editing is completed. The "special color plate" is a plate for printing with an ink other than ordinary CMYK, and for example, (1) black and gold red "gold red" in two-color printing, (2) gold, silver, (3) When expressing the color of a woman's lipstick in a photograph in a colorful way, use it as a separate plate for the lipstick, adding it to normal CMYK, and stacking ink on it. or,
“Trapping” specifically means stacking little by little in order to prevent blank areas due to misalignment of printing when they are arranged by removing hair. Server workstation 30
The data that has undergone the plate-making / editing process is processed by the galley printer 20 for galley output and PDL (Page Description).
n Language; eg PostScript)
The output is made, or the film is outputted by the film printer 60.

With such a configuration, the area extraction method of the present invention will be described with reference to the flowchart of FIG. Here, the CRT 303 of the editing workstation 10 displays editing target images G1 and G2 as shown in FIG. 9, and the following description will be made assuming that the image G1 in the editing target images G1 and G2 is cut out as the plate collecting operation. To do. An example using the mouse 305 as the pointing means (keyboard 304, mouse 305, digitizer 306, trackball, joystick, etc.) will be described. The process relating to the present invention is performed by the plate-collecting module 360 shown in FIG.

The operator first sets the parameters relating to the area extraction. Parameters related to region extraction include “observation width” and “observation resolution (threshold, smoothness)”, which will be described later. If you want to change these setting values,
The "option" menu (not shown) in the function selection menu displayed on the screen is designated by the mouse 305. An input window opens in a predetermined portion on the screen according to an instruction from the "option" menu, and reference values of the respective parameters are displayed in the input window 6 as shown in FIG. Here, the "length" shown in the input window 6 is the "width W (= observation width) of the band-shaped polygonal line 2" for designating the contour portion 1 of the image G1 to be cut out. Threshold ”
Is a threshold regarding the edge strength of the line 1 to be cut off. Further, “smoothness” is a value related to the detection interval of the edge points (shift interval of the observation area 7 described later). If the smoothness setting value is increased, the edge points are detected at a narrower interval. It has become. The operator arbitrarily sets the observation width and the observation resolution according to the shape, color shade, etc. of the image to be cut. In addition,
Press "Cancel" in the input window 6 to redisplay the reference value (or the latest set value), press "OK" to close the window, and display the set value in the plate collection module 360. It is passed to the CT boundary processing section. When the operator does not set the observation width and the observation resolution, the area extraction processing is performed using the preset reference value (step S10).

After setting the parameters relating to the area extraction, the operator uses the mouse 305 to press the "CT image boundary recognition icon" (not shown) in the function selection menu displayed on the screen. . Then, using the mouse 305, the outer frame region is instructed and input in a band shape so as to include the contour portion (line to be cut) 1 of the image G1 in FIG. Here, the instruction of the line 1 to be cut indicates a main point outside the line 1 to be cut, such as points P1, P2, ..., Pn as shown in FIG. When each point is designated, on the screen, as shown in the figure, an outline polygonal line 2a connecting the points P1, P2, ..., Pn and a constant width W with respect to the outline polygonal line 2a.
A strip-shaped polygonal line 2 (hereinafter referred to as a general polygonal polygonal line) including an inner polygonal polygonal line 2b having (a value set by an operator or a reference value) is displayed. Here, if it is desired to correct the position of the point or to add a point, those instructions are given, and when the instruction of the line 1 to be cut is finished, the mouse 305 is double-clicked (step S
11).

By this operation, the automatic recognition processing of the extraction area is started. In the CT boundary processing unit in the plate-collecting module 360, the difference between the pixel of the color image G1 in the outer frame area and the adjacent pixel in the color space (in the case of a monochrome image, the density of the pixel with the adjacent pixel in the continuous tone image). Difference),
A partial image that connects the pixels showing the maximum difference to form the outline,
The extracted area in the original image. Explaining in detail with reference to the drawing, the CT boundary processing unit first prepares an observation cross buffer 3 as shown in FIG. 10 on the memory according to the observation width W set in step S10. Then, the center of the observation region 7 (see FIG. 9) corresponding to the observation cross buffer 3 is set so as to come to the center of the outline polygonal line 2, and the coordinate value of the observation region 7 at the observation start point is calculated (step S
12). Then, the pixel value in the observation area 7 is set in the observation buffer 3 (step S13). At that time, it is determined whether or not the inclination angle of the outline polygonal line 2 is larger than 45 degrees with respect to the X-axis, and when the inclination of the outline polygonal line 2 is less than 45 degrees, as shown in FIG. In the buffer of the observation buffer 3 in the Y-axis direction, the coordinate value (Ea in the figure) that maximizes the distance between adjacent pixel values is detected as an edge point. Further, when the inclination of the outline polygonal line 2 is 45 degrees or more, the distance between adjacent pixel values becomes maximum in the buffer of the observation buffer 3 in the X-axis direction as shown in FIG. The coordinate value is detected as an edge point (Eb in the figure).

Here, when the distance between adjacent pixel values is d, the adjacent distance d is calculated by the following expression 1 in the case of a color image and by the following expression 2 in the case of a continuous tone image. However, when the difference from the background (for example, the image G2 in FIG. 9) is unclear and the pixel indicating the maximum difference cannot be detected, that is, when the calculated maximum difference is less than the “threshold”,
As shown in FIG. 11C, the centers of adjacent pixels (Ec in the figure) are set as edge points. That is, when there is no edge candidate point in the local area, the boundary of the area can be set by setting the center of the outline polygonal line designated and input by the operator as the edge point even if the edge cannot be detected. This eliminates the need for arithmetic processing such as a two-dimensional filter (step S14).

[0022]

[Equation 1] However, r1, g1, b1, r2, g3, and b3 are the values of the red (R), green (G), and blue (B) components of adjacent pixels.

[Equation 2] However, t1 and t2 are the density values of adjacent pixels.

When the edge points in the observation area 7 are detected, it is judged whether or not the edge point detection processing up to the observation end point is completed (step S15). If not, the coordinate values of the observation area 7 are determined. To the next observation point, the process returns to step S12 (step S16), and the edge point detection process is repeated. In this way, the pixel value is set in the observation cross buffer 3 while traveling in the observation region 7 along the center line of the outline polygonal line 2, and the pixel having the maximum difference is sequentially detected as an edge point. The traveling interval at that time is determined based on the value of "smoothness" set in step S1. In step S15, if the edge point detection processing from the observation start point to the end point has been completed, the CT boundary processing unit in the stencil module 360, as shown in FIG. 12, obtains the obtained edge point (e1, e
, ...) are connected by an approximate curve by a known technique such as the least square method, the partial image having the connected edge line 4 as the contour line is set as the extraction region 5, and the region extraction processing is completed.

In the above-described embodiment, the area extraction method in the image processing system having the workstation as a constituent element has been described as an example.
As a matter of course, it can be applied to a personal computer, a word processor, or the like.

[0025]

As described above, according to the area extraction method of the present invention, an arbitrary area in an unspecified color image (or continuous tone image) can be extracted only by roughly specifying it. Can be significantly reduced. Moreover, since the calculation process for detecting the contour line of the extraction region is extremely simple, the region extraction process can be performed at high speed.

[Brief description of drawings]

FIG. 1 is a system configuration diagram showing an example of a hardware configuration of an image processing system which is a premise of the present invention.

FIG. 2 is a system configuration diagram showing another example of the hardware configuration of the image processing system on which the present invention is based.

FIG. 3 is a block diagram showing an example of a hardware configuration of an editing / server workstation.

FIG. 4 is a block diagram showing a configuration example of the entire system of FIG.

FIG. 5 is a block diagram showing an example of a detailed software configuration of a server workstation.

FIG. 6 is a flowchart showing the flow of the entire editing work.

FIG. 7 is a diagram showing a relationship between input target data and a software process.

FIG. 8 is a flowchart showing an outline of a region extraction method of the present invention.

FIG. 9 is a first diagram for explaining the method of the present invention.

FIG. 10 is a second diagram for explaining the method of the present invention.

FIG. 11 is a third diagram for explaining the method of the present invention.

FIG. 12 is a fourth diagram for explaining the method of the present invention.

FIG. 13 is a fifth diagram for explaining the method of the present invention.

[Explanation of symbols]

 1 line to be cut 2 outline polygonal line 2a outer polygonal line 2b inner polygonal line 3 observation cross-buffer 4 outline of extraction area 5 extraction area 6 input window 7 observation area 10, 30, 30A workstation 302 hard disk 303 CRT 304 keyboard 305 mouse 306 Digitizer 307 Floppy disk 311 Data disk 360 Plate collection module 360

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Office reference number FI technical display location 9061-5H G06F 15/70 335 A

Claims (4)

[Claims] 1. A storage means for storing image data of a color image, a display means for displaying the color image, and a pointing means for indicating and inputting arbitrary coordinates in the displayed color image. In the area extraction method of the image processing device, the original image stored in the storage means is displayed on the display means, and the outer frame area is instructed and input by the pointing means so as to include the contour portion being displayed, The difference between the pixel of the original image in the outer frame area and the adjacent pixel in the color space is obtained, and the partial image that connects the pixels showing the maximum difference to form the contour line is set as the extraction area in the original image. A region extraction method characterized by the above. 2. Storage means for storing image data of a continuous tone image, display means for displaying the continuous tone image, and designation and input of arbitrary coordinates in the displayed continuous tone image. In an area extraction method for an image processing apparatus including pointing means, the original image stored in the storage means is displayed on the display means, and the outer frame area is defined by the pointing means so as to include the contour portion being displayed. A band-shaped instruction input is performed to obtain a difference in density between adjacent pixels with respect to the pixels of the original image in the outer frame area, and a partial image that connects the pixels showing the maximum difference to form a contour line is defined as an extraction area in the original image. An area extraction method characterized in that 3. The method according to claim 1, wherein the outer frame region is scanned in a cross pattern to calculate the difference.
The area extraction method described in. 4. The area extraction method according to claim 3, wherein when the pixel showing the maximum difference cannot be detected, the center line of the outer frame area is set as the contour line.