JPS61208175A - Detection system for area - Google Patents

Abstract

PURPOSE:To detect color areas of a color image and to set color areas easily at a high speed by performing area detection while expanding an area of color similar to that of a point specified by an operator. CONSTITUTION:When the operator monitors an image displayed on a CRT 6 and indicates one or plural points on a tablet 1, RGB values corresponding to colors of the indicated points are assigned in a bit sequence. Similar colors have RGB values which have same high-order digit bits of respective bit sequence and the different low-order digit bits. For the purpose, the low-order digit bits are masked and RGB values of all adjacent points are read in and then compared with the high-order digit bits to a determine an area having a color similar to these of the indicated points. Points of coincident color data are fetched and then the area to be compared is shifted to the high-order digit bits, so that a comparison is further made.

Description

[Detailed Description of the Invention] [Summary] This detects the color area in a color image displayed on a display means such as a CRT, and expands the color area similar to the point specified by the operator by color mask processing. To perform area detection and quickly and easily set color areas.

[Industrial application field]

The present invention relates to an area detection method for detecting a color area in a color image displayed on a display means such as a CRT, and more specifically, the present invention relates to an area detection method for detecting a color area in a color image displayed on a display means such as a CRT. Regarding the method.

In order to cut out a region having a specific color from an image having many colors, such as a natural image, a desired region is traced using an input device such as a tablet or an input pen. However, since processing complex graphics requires manpower and time, it is desired to provide an area detection method that is easy to operate and capable of high-speed processing.

[Conventional technology]

Region detection of a color image is performed by tracing and cutting out the outline of a desired region using an input/pointing device such as a mouse, an input pen, or a tablet. On the other hand, it has been considered to automatically cut out a closed region containing one color or a color similar to this, but setting the region is difficult and has not been implemented yet.

[Problem that the invention seeks to solve]

A method of manually cutting out an area using an input/pointing device can accurately cut out a desired area, but it takes time to cut out areas having the same color in a complex figure such as a natural image. Additionally, a method of automatically finding closed regions containing the same or similar colors has been considered, but it has not been implemented because it is difficult to set the color regions.

An object of the present invention is to provide an interactive area detection method that can quickly and easily set or detect color areas.

[Means for solving problems]

FIG. 1 is a block diagram showing an example of a hardware configuration suitable for implementing the area detection method of the present invention, and FIG. 2 is a flowchart thereof. In FIG. 1, 1 is a tablet for supplying instruction data to a bus 2, 3 is a main memory for storing programs or data, and 4 is a processor (C
PU), 5 is a frame memory that stores images to be displayed on the CRT 6.

[Operation] When the operator specifies one or more points from the tablet 1 while monitoring a figure, natural image, etc. displayed on the CRT 6, the RGB values corresponding to the color of the specified point are read out, and then A reference value with lower bits masked is created 0
Addresses with values within the reference values are then scanned to determine areas having a similar color to the indicated point. Here, when the operator instructs to expand the area, a reference value is set using the mask bits shifted to higher order, and a comparison is performed using this reference value to detect the expanded area. These processing operations continue until the operator instructs them to end.

How to obtain the reference value will be explained with reference to FIG. 3 showing a bit string and FIG. 4 showing a mapping of this to a three-dimensional space. The third vIJ is a diagram showing a bit string (color data) for expressing a color by allocating 8 bits to each of the RGB values conventionally used for color processing in electronic needle counters. Referring to the figure, it will be understood that similar colors have RGB values in which the higher bits are the same and the lower bits are different in each bit string. FIG. 4 is an explanatory diagram conceptually showing a point C in space indicated by each RGB component shown in FIG. 3, and an area d expanded when the lower bits are masked. In addition, RG
If the mask is applied only to the second-order component in the B value, the area d
becomes a plane.

〔Example〕

FIG. 2 is a flow diagram of the area detection method according to the present invention.

When the operator specifies one or more points on the image containing many slightly different colors using the tablet 1 in step 10, RGB reference values are created for each of the specified points in step 20, and then in step 30.
A process is performed to find all points of the same color adjacent to the specified point. This process continues until there are no adjacent points of the same color. As a result, areas of the same color are determined and displayed on a display means such as a CRT 6. Next, in step 40, when the operator inputs from the tablet 1 or the like that the area displayed on the display means such as the CRT 6 is less than the desired area, in step 50, the lower bits (color approximation display bits) of the color data of the designated point are input. The next masked reference data is generated. That is, when n (≧1) bits are given to each of RGB, the lower m (n
≧m≧1) Data with masked bits is generated.

Further, in step 60, after the color data of all points adjacent to the surface area are read, a comparison with the upper bits is performed. When points of matching color data in one adjacent area are captured, the area to be compared is expanded and further comparison is performed. As a result, the lower bits are masked and a new surface area set by the upper bits is determined and displayed on display means such as the CRT 6. If the operator desires further expansion in step 70, the mask bits are shifted upwards and processing continues. In this way, the operator can detect the desired area. Note that according to the present invention, it is clear that the operator can specify points of arbitrary different colors at the same time, and the present invention also includes a method of specifying a range that includes any of the colors of these points. can be adopted without departing from the main idea. Further, in this embodiment, when performing area detection processing, mask bits are applied to each of RGB at the same time for comparison and display, but mask bits may be applied to each of these in sequence for comparison and display. , it is also possible to perform comparison and display by applying independent mask bits to RGB.

〔Effect of the invention〕

As described above, according to the present invention, it is possible to quickly and easily detect the area in an interactive manner while expanding the area of color similar to the point specified by the operator, and the area can be configured with any color. This has the advantage that an image can be quickly and easily divided into regions of any desired number of colors.

[Brief explanation of drawings]

Figure 1 is a block diagram showing an example of the hardware configuration, Figure 2 is a flowchart for the hardware configuration shown in Figure 1, Figure 3 is an explanatory diagram of a bit string, and Figure 4 is the concept of mapping a bit string to a three-dimensional space. Figure, 1st
In the figure, 1 is a double-left, 2 is a bus, 3 is a main memory, and 4 is a CPU. 5 is a frame memory, and 6 is a CRT. Figure 1

Claims (2)

[Claims] (1) In an image processing system, a first means for specifying an arbitrary point having color data for an area to be detected, and detecting an area having color data from which a comparison with the specified point can be obtained; a second means for displaying the detected and displayed area together with the background area; and a second means for masking the color approximation display bits in the color data of the specified point and using that color data when the detected and displayed area is less than the desired area. and third means for causing the second means to perform the detection and display. (2) The area detection method according to claim 1, wherein the third means includes means for changing the mask applied to the color approximation display bits.