JPS6346573A - Color image processor - Google Patents

Abstract

PURPOSE:To regard various types of hues or undesired areas in a background as backgrounds respectively and to satisfactorily segment and synthesize images, by using an area designating memory, a background color memory, an image processing controller, etc. CONSTITUTION:Areas to be segmented and deleted are designated manually and roughly to an area designating memory 60. When the background colors are deleted, a background color is designated by having inputs of several key points by means of a cursor. The hue of a designated point is defined as a background color after setting the vertical width on the R-G-B and setting up the bit of a background color memory 80. At the same time, the data set between frame memories A20 and B70 can be freely written and transferred by an image processing controller 100. Then the memory 60 controls the planar direction of a screen; while the memory 80 controls the hue depth direction of each point. Thus some points are regarded as backgrounds even though the background of a synthesized image includes various hues. Furthermore, the undesired areas in the backgrounds are also designated and regarded as backgrounds. Thus it is possible to satisfactorily segment and synthesize the images.

Description

DETAILED DESCRIPTION OF THE INVENTION Technical Field The present invention relates to a color image processing device that cuts out a partial area of an original color image.

<Prior Art> Various processes have been performed to cut out a part of an original color image and combine it with other images.

In this cutting, the key point is the separation of the cutting area from the background. Various separation methods have been proposed and put into practical use. For example, there is a method known as the chroma-ki method that recognizes a specific hue as the background color, a method that uses edge detection to find the outline of the object, or a method that manually cuts out the object using some kind of pointing device (such as a light ben or digitizer). There is. The former chroma system has the disadvantage that when there is the same hue in the object, it is recognized as the background, and the latter method, which calculates the contour, may detect unnecessary edges during edge detection, or it may be too difficult to calculate manually. The drawback was that it was time-consuming.

(Purpose) The present invention eliminates the above-mentioned drawbacks, and at the same time makes it possible to recognize various hues as a background, which was impossible with the chroma system, and also to recognize unnecessary areas in the background as a background. The purpose of this paper is to provide a color image processing device that is capable of color image processing.

Embodiment FIG. 1 shows an embodiment of the present invention, in which two memories, an area designation memory and a background color memory, are added to the conventional frame memory and image processing CPU. To give a brief overview below, the manually inputted image signal is converted into a digital signal (No. 3) by the A/D converter 10,
20 is input. The image stored in the frame memory A 20 and the area specified in the area specification memory 60 are overlapped, and are constantly sent to the converter 40 by the timing controller 30 and displayed on the monitor display 50. These image data are transferred via video bus VB.

On the other hand, another image to be synthesized is stored in the frame memory B70, and the data between the frame memories A20 and B20 can be freely read, written, and read by the image processing controller 100 via the system bus SB. Transfer is now possible. Furthermore, a background color memory 80 is connected on the system bus SB.

The frame memory A20.870 is a full-color image memory with 640 dots horizontally x 480 dots vertically, 8 bits each for R, G, and B, and the area specifying memory 60 is also a memory of the same vertical size and 8 bits deep. Of these 8 bits, 1 bit is used to designate the compositing area, 1 bit is used to designate the background area, and the others are used for work.

Further, a pointing device 90 such as a digitizer tablet is connected to the image processing controller 100, and the current position is displayed on the monitor display 50.

Now, assume that the frame memory A20 contains a portrait as shown in FIG. 2, and the frame memory B70 contains a landscape as shown in FIG. The background of the portrait is a wall, but there is some shading, so it is not a blank space.

Also suppose that the clothes contain a circle pattern that seems to have the same hue as the wall color. I would like to combine the images in frame memories A and B into an image as shown in FIG. To do this, you can cut out only the person you want to combine, but
It is difficult to trace the outline faithfully, and it is impossible to trace a thin part like a single hair. First, cut out the rough part inside the part that is likely to be confused with the background color or the part you want to combine. The cutting is performed by using a pointing device 90 such as a digitizer or tablet, and displaying a cursor on screen A, and filling in the area. This area is defined as the compositing area.

FIG. 5 shows the cutout portion of the screen displayed on the monitor with horizontal lines. Conversely, if there is a part in Figure 2 that you want to remove, specify a large part outside of it in the same way. Make this area the background area. The composite area and background area are displayed in different colors on the monitor, but only the area specification memory is changed, and the original image is not modified.

As described above, the area to be cut out and the area to be removed are roughly specified manually in the area specification memory.

The next step is to remove the background color.To specify the background color, input several representative points with the cursor, set the hue of the specified point to the upper and lower middle of R-G-B, and then Set the memory bits as the background color. For example, R-G-B is 50-
When the value is 100-200, if the upper and lower middle are ±10, then 0
- Set bits in the R-G-B area that satisfy all expressions 40≦R≦60 □ ■ 90≦G≦110 □ ■ 190≦B≦210 □ ■.

Repeat this for the number of points you manually achieved.

The size of this memory is 2 if each R, G, and B are 8 bits.
'x2'x26=2''=16 Mbits, which is a huge amount, but if you consider having upper and lower middles, 5 bits each is considered sufficient, and in that case, 25 x 2 S x 2
This can be realized with a small capacity RAM of 32K bits.As described above, the area specification memory controls the plane direction of the screen, and the background color memory controls the depth direction of the hue of each point. .

In actual compositing, the image processing controller scans the screen point by point in the horizontal and vertical directions, and when the area designation memory specifies the compositing area, it replaces the pixels on the screen with screen B, and then scans the screen point by point in the horizontal and vertical directions. If neither is specified, do nothing; if neither is specified, refer to the background color memory; if the background bit is not set, change the pixels of screen A to screen B.
6. As explained above, according to this embodiment, even if the background of the image to be synthesized is not uniform, such as shading, several points can be pointed and regarded as the background. Even if the color has the same hue as the background color, by setting a mask bit, it will not be confused with the background, and unnecessary areas in the background can be considered as the background by specifying them in advance. , it is possible to cut out and synthesize good images.

<Effects> As described above, according to the present invention, it is possible to cut out the image in detail.

[Brief explanation of the drawing]

Figure 1 is a processing block diagram of this embodiment, Figure 2 is a diagram showing images in frame memory A, and Figure 3 is a diagram showing images in frame memory B.
FIG. 4 is a diagram showing a composite image, and FIG. 5 is a diagram showing a cut-out portion of FIG. 2.

Claims (1)

[Claims] It is characterized by comprising a setting means for setting an area of the original color image, a specifying means for specifying a plurality of background colors outside the area, and a means for cutting out a part of the original image using the outputs of the setting means and the specifying means. Color image processing device.