JPH07322054A - Area setting method - Google Patents

Abstract

PURPOSE:To easily and speedily set the area of only a subject part by designating a rough area including a desired subject in an image displayed in CRT and executing a necessary subject recognition processing on the image of the designated area. CONSTITUTION:An image pickup signal by a CCD camera 4 is reproduced by CPU 6 and the like and the reproduced image is displayed on CRT 14 of a console 13. When the subject 21B and the like in the image 20 is selected and the rough area 30 is designated through a mouse 16, the subject recognition processing of the subject 21B such as an outline recognition processing and a skin color recognition processing is executed as against the area 30 through CPU 6, a processing condition storage part 8 and memories 9-11. Then, a subject image area is easily and speedily set by following a subject outline without designating an image area.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an area setting method for setting a processing area for performing image processing on a photographic image.

[0002]

2. Description of the Related Art Conventionally, there are various methods for obtaining a desired image by performing image processing such as gradation correction, color correction, and edge enhancement on an image signal representing an original image read by an image scanner, a video camera or the like. Although proposed, if the entire image is uniformly processed without specifying the processing range for the original image, there will be a portion that is appropriately processed and a portion that is not appropriately processed. .

Therefore, for example, it is read by an image scanner or the like on a personal computer or a workstation,
A method has been proposed in which a processing area is specified for an original image displayed on a CRT by using a light pen or the like, and desired image processing is performed on that portion. In addition, when specifying the processing area in this way, a step may occur in the image at the boundary portion of the area, so that the area setting that can prevent unnaturalness in this strengthened portion Has been proposed (Japanese Patent Laid-Open No. 5-250462, etc.).

Further, in the field of an apparatus for reading a document and processing the document for recording, only a certain region of the document is recorded, or a certain region in each of two pages of the document is recorded on one page of recording paper. You may want to perform trimming or overlay. Therefore, a method is used to set an area in the original with a specific color and recognize this area when reading the original, or to set an area in the original with a loop and use this loop as an area for processing. Recognition methods have been proposed (JP-B-6-2363, JP-A-57-194670, JP-A-57-196660, JP-A-57-203372, etc.).

On the other hand, also in the field of photographic images, for example, in the case of a photographic image in which a person is reflected, an area including a person is set on the image, and the density and gradation of the image in this area are set. There are cases in which a photographic image is created in accordance with the user's wishes by changing, erasing or trimming. In such a case, the user specifies the subject to be processed by looking at the printed positive image, and specifies the region by instructing the operator of the camera store having the image processing system to this region. The operator then puts a positive image on the reading device and prints this image on C
A region including only the desired subject to be displayed on the RT and using the mouse or the like on the CRT to be processed as instructed by the user is set, and the image in this region is instructed by the user (for example, trimming, Density correction, gradation correction, red-eye correction, etc.). As a method of setting this area, for example, a method of automatically drawing a boundary using a color difference known as chroma key on a personal computer or a workstation, and tracing the boundary line using the mouse described above A method is used in which a boundary portion is roughly selected and a boundary line is created by utilizing a color difference and a brightness difference in the boundary portion.

[0006]

As described above, in order to perform various image processings on an image, it is necessary to set an area in the image. When setting this area, the desired processing should be performed. It is preferable to precisely trace the contour of the subject to set the region. However, it takes a very skilled technique to accurately trace the contour of the subject and set the area, and it also takes a long time to set the area. Therefore, when setting a region, a rough region including a portion other than the subject near the desired subject to be processed is often set.

However, if a rough area including a portion other than the subject is set and the above-described image processing is performed, an unnatural boundary is formed between the processed area and other areas. The image becomes unnatural.

In view of the above circumstances, the present invention roughly sets an area including a predetermined subject to be subjected to image processing, and can set only a part of the predetermined subject as an area to be subjected to image processing. It is an object to provide an area setting method.

[0009]

A region setting method according to the present invention sets a region of a desired subject in the image in order to perform predetermined image processing only on the desired subject in the image as described above. In this method, a desired subject is selectively specified by selectively including only a desired subject and specifying a region including a portion near the subject in the image, and subjecting the image within the designated region to subject recognition processing. Is set in the image as an area of a desired subject.

Here, the subject recognition process may be a process of recognizing a desired contour line of a subject in an image in a designated region, or a process of performing skin color recognition in an image in the designated region. preferable. In this case, the area surrounded by the contour lines recognized by the contour line recognition processing or the skin color area recognized by the skin color recognition is set as the desired subject area.

[0011]

The area setting method according to the present invention selectively specifies only a desired subject to be image-processed, and only specifies a rough area including the area near the subject in the image. The subject recognition processing is performed on the region to set a region corresponding to a desired subject in the image. Therefore, it is possible to set only the region corresponding to the subject to be processed in a short time without accurately tracing the contour of the subject when setting the region on the image. Therefore, since the boundary between the processed area and the other area is located at the contour portion of the desired subject, an unnatural boundary between the processed area and the other area may be created. To be prevented. Also,
Even when image processing is performed within this area, it is not necessary to perform processing on a portion other than the desired subject.

When the flesh color recognition processing is performed as the above-mentioned subject recognition processing, when the desired subject is a human, the area corresponding to this desired subject can be satisfactorily detected, and the contour line When the recognition processing is performed, for example, when a subject surrounded by a closed region such as a human face is set as a desired subject, the region corresponding to the desired subject can be favorably detected.

[0013]

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

FIG. 1 is a diagram showing an outline of an apparatus for carrying out the area setting method according to the present invention.

As shown in FIG. 1, an area setting device 1 according to an embodiment of the present invention reproduces a negative image on which an original image is recorded and a visible image reproduced on a recording medium such as a photosensitive material based on the negative film. A CCD camera 4 for reading an image signal from a medium 2 such as a positive image through an RGB switching filter 3, and an A / D conversion means 5 for A / D converting an image signal representing an image output from the CCD camera 4. And CPU6
A ROM 7; a processing condition storage unit 8; a memory 9 for storing a read image signal; and a memory 10 as an area storage unit for storing an image in a set area. A memory 11 as a processed image storage unit for storing an image signal subjected to image processing as described later.
A printer 12 for recording the processed image signal on the photosensitive material as a visible image, and a console 13, and these respective components are interconnected via a bus 17.

The CPU 6 controls the entire apparatus 1, and R
The OM 7 stores a program for defining the processing procedure in the CPU 6. Further, the processing condition storage unit 8
Stores the processing conditions set by the console 13 for the images in the area. The memory 9 is the RGB filter 4
The image signals corresponding to the R (red), G (green), and B (blue) color images separated by the CCD camera 4 are stored. The memory 10 as an area storage unit stores an area set to perform desired image processing on the original image as an area signal created as described later.

The console 13 is provided with a CRT 14, a keyboard 15 and a mask 16 and is used to input an instruction for performing processing as will be described later on an original image or a processed image in an area.

First, as shown in FIG. 2, the CCD camera 4 reads a reproduced image 20 reproduced on a recording medium such as a photosensitive material. In this reproduced image 20, three persons 21A, 21B and 21C are shown as shown in FIG. Then, it is assumed that the user desires a process for changing the density of the face of the person 21B.

This reading is performed based on the flow chart shown in FIG. That is, the reproduced image 20 is read by the CCD camera 4 with the RGB switching filter 3 set to R (steps S1 and S2). The read original image is converted into a digital signal by the A / D converter 5 and then stored in the memory 9 via the bus 17 (step S
3). Next, the operations of steps S2 to S3 are repeated with the RGB switching filter 3 set to G and B, and the image signals S ₁ representing the respective reproduced images 20 are stored in the memory 9 (step S4).

Based on the image signal S ₁ obtained in this way, a region corresponding to a desired subject is set as described below.

FIG. 4 is a flow chart showing the steps of area setting in the area setting method according to the embodiment of the present invention.

As shown in FIG. 4, the image signal S ₁ is first input to the CRT 14 of the console 13 and displayed as a visible image here (step S11). Next, the operator selectively includes only a desired subject with the mouse 16 or the keyboard 15 and designates a region including a portion near the subject on the CRT 14 (step S12). Here, in this embodiment, as the desired subjects to be subjected to image processing later, the three persons shown in FIG.
The person 21B is selected from 21A, 21B, and 21C. Therefore, as shown in FIG. 5, only a person 21B is selectively included on the CRT 14, and a region 22 (a hatched portion in FIG. 5) near the person 21B is included in the area.
30 is specified.

Then, subject recognition processing is performed only on the image in the area 30 (step S13). In this embodiment, since the desired subject in the area 30 is a person, the skin color recognition process is performed.

As a skin color recognition method, first, for the image signal S ₁ ′ representing the image in the region 30 reproduced on the CRT 14, the combination of the densities of the B, G and R components of this image signal S ₁ ′ is used as the axis. The three-dimensional coordinate system is set, and the region representing the skin color in this coordinate system is defined by a solid having an elliptical or circular cross section. Then, when the image signal value of each of the B, G, and R components in the image signal carried by the pixel in the region 30 reproduced on the CRT 14 is included in this solid, the pixel carrying the image signal The part of is recognized as a skin color. This flesh color recognition method is specifically described in JP-A-52-156624.
No. 60-217353.

When the flesh color is recognized in this way, the region 31 corresponding to the face of the person 21B in the image in the region 30 is recognized as the flesh color as shown in FIG.

Then, the area 31 recognized as the skin color is set in the reproduced image 20 displayed on the CRT 14 as an area to be subjected to image processing desired by the user (step S14). The image signal S ₁ ″ representing the image corresponding to the area 31 thus set is temporarily stored in the memory 10.

As described above, according to the area setting method of the present invention, the face of the desired subject 21B to be processed is roughly designated in the reproduced image 20 reproduced on the CRT 14 so that the desired area of the desired subject 21B can be obtained. Area corresponding to the face part of
Only 31 can be set in the reproduced image 20.

With respect to the image signal S ₁ ″ in the area 31 detected by setting the area 31 in this way, processing specified by the user, for example, processing such as density enhancement and gradation change is performed. Input from the mouse 16 and the keyboard 15 of the console 13, and the CPU 6 performs image processing on the image signal S ₁ ″ according to the processing conditions stored in the processing condition storage unit 8 according to these inputs (step S15).

[0029] "When the image processing is performed on, then the image processing is image signals S ₁ was made" Thus the image signals S ₁ in the synthesis of the image represented by the image signals S ₁ is Nasaru. That is, as shown in FIG. 7, the image 34 represented by the image signal S ₁ ″ subjected to image processing and the image 32 other than that are combined.

The image signal S ₂ representing the image thus synthesized is once stored in the memory 11, and then the image signal S ₂ is input to the printer 12 so that the image represented by the image signal S ₂ is a photosensitive material or the like. Is reproduced as a visible image on the recording medium.

As described above, by using the area setting method according to the present invention, it is only necessary to roughly specify the area including only the desired subject to be subjected to the image processing in the reproduced image, and only the desired subject is imaged. Since the area to be processed is set, it is not necessary to accurately trace the contour of the desired object while observing the CRT 14, and it is possible to set the area to be subjected to image processing only to the desired object in a short time. it can. Further, since the boundary between the image-processed area and the other area is located at the boundary between the person 21 and the other area, there is an unnaturalness between the processed area and the other area. Boundaries are prevented. Further, even if the area is roughly specified, it is not necessary to perform the image processing on the extra area other than the desired subject, so that the calculation time for the image processing can be shortened.

In the embodiment described above, the skin color recognition is performed in step S13 of the subject recognition process shown in FIG. 4, but the invention is not limited to this, and the process of recognizing the contour line of the subject is performed. May be applied.

When the image signal representing the image in the area 30 is differentiated as shown in FIG. 8A, the differential value, for example, I-
A cross section taken along line I is shown in FIG. Therefore,
In this way, a pixel having an image signal value such that the differential value shows a peak and the peak is continuous is defined as a contour line, and an area 33 surrounded by the contour line corresponds to a desired subject to be image-processed. It may be recognized as a region to be operated.

Further, in the above-mentioned embodiment, the reproduced image recorded on the recording medium such as the photosensitive material is read by the CCD camera 4 to set the area including only the desired subject in the image obtained. However, the present invention is not limited to this. For example, an image signal obtained by reading the image recorded on the negative film with the CCD camera 4 is C
The area may be displayed on the RT and the area may be set, or an image signal representing an image such as a photo CD may be directly displayed on the CRT and the area may be set on the CRT. Further, an area including a desired subject is directly written in a reproduced image recorded on a recording medium such as a photosensitive material, and the reproduced image is read by the CCD camera 4 together with the area to include the desired subject on the CRT. The subject recognition process may be performed immediately without specifying the area.

In the subject recognition processing described above,
Regarding the object for which the contour recognition processing is performed, the desired subject is not limited to a person, and any subject including a closed region may be selected as the desired subject to be subjected to image processing. It's good.

[0036]

As described in detail above, the area setting method according to the present invention can roughly set an area including only a desired subject to be image-processed in an image.
Since the area including only the desired subject can be set in the image, it is not necessary to accurately trace the contour of the desired subject in the image when setting the area,
It is possible to set an area to be subjected to image processing in a short time. In addition, since the boundary between the processed region and the other regions is located at the outline of the desired subject,
It is possible to prevent an unnatural boundary between the processed area and the other areas. Further, even if the area is roughly set, it is not necessary to perform the image processing on the extra area other than the desired subject, so that the calculation time for the image processing can be shortened.

[Brief description of drawings]

FIG. 1 is a schematic diagram of an apparatus for carrying out a region setting method according to the present invention.

FIG. 2 is a diagram showing a reproduced image including a desired subject reproduced on a recording medium such as a photosensitive material.

FIG. 3 is a flowchart for explaining reading of an image signal from a reproduced image.

FIG. 4 is a flowchart for explaining a region setting method according to the present invention.

FIG. 5 is a diagram showing a state in which a reproduced image selectively includes only a desired subject and a region including a portion near the subject is designated.

FIG. 6 is a diagram showing a state in which an area corresponding to only a desired subject is set.

FIG. 7 is a diagram for explaining composition of an image corresponding to a desired area that has been subjected to image processing and an image other than that.

FIG. 8 is a diagram for explaining recognition of a contour line.

[Explanation of symbols]

 1 area setting device 2 medium 3 RGB switching filter 4 CCD camera 5 A / D conversion means 6 CPU 7 ROM 8 processing condition storage unit 9, 10, 11 memory 12 printer 13 console 17 bus 20 replay image 21A, 21B, 21C person 30 Rough area 31 Area to be subjected to image processing

─────────────────────────────────────────────────── ─── Continuation of the front page (51) Int.Cl. ⁶ Identification code Office reference number FI Technical display location H04N 1/46 7459-5L G06F 15/70 330 Z H04N 1/46 Z

Claims (3)

[Claims] 1. A method for setting a region of a desired subject in the image for performing predetermined image processing only on the desired subject in an image, the method comprising selectively including only the desired subject. A region including a portion near the subject is designated in the image, and subject recognition processing is performed on the image portion in the designated region, so that a region corresponding to the desired subject is set as the region of the desired subject. An area setting method characterized by setting in an image. 2. The subject recognition process is a process of recognizing a contour line of the subject for an image in the designated region, wherein a region surrounded by the contour line is set as a region of the desired subject. The area setting method according to claim 1, wherein the area setting method is provided. 3. The subject recognition process is a process of performing flesh color recognition on an image in the designated region, and the flesh color region recognized by the flesh color recognition is set as the desired subject region. The area setting method according to claim 1.