JPH11185048A - Method and device processing image - Google Patents

Abstract

PROBLEM TO BE SOLVED: To prepare a picture of a face as imaged by performing morphing processing to a basic face best matched with the image and reference face having features in parts, and providing an image of a face existing between these faces. SOLUTION: Through a face selecting means 111, an operator selects a basic face closest to an image and a reference face having features in parts to be worn at one part of a face or body. Data on the image are transferred from a data library 112 to a buffer memory 113 and displayed on a display 115. The operator watches the display 115 and selects one basic face matched with the contour or image of the face and one reference face having the features in the parts. A morphing processing means 117 prepares a picture between the basic face and the reference face, and a prepared basic face is printed as a montage by a printer 120 as it is.

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image processing method and an image processing apparatus for producing a montage photograph used for, for example, a scientific investigation. In this specification, "image" has a broad meaning including a photograph itself, a data-processed thing, and the like. In this specification, “parts” refers to, for example, eyes,
In addition to the parts such as the nose and mouth, it refers to a broad meaning including things worn by a person, such as glasses and clothes.

[0002]

2. Description of the Related Art A conventional montage photograph classifies a large number of photographs related to a crime or the like, and selects a photograph determined to be the closest from the above photographs based on an impression of a first discoverer with respect to a criminal. Is used. In addition, the above selected photograph was made closer to the real thing by replacing some of the parts. Another conventional montage photograph was created as a portrait by a person with special skills based on information and photographs of the first discoverer of the criminal. Further, in recent years, due to the development of computers, montage photography has a library of a large number of photographs stored in a computer, and the first discoverer of the culprit selects photographs that are relatively impressive, and It is produced by performing image processing such as replacing parts consisting of eyes, nose, mouth and the like. Further, after the above parts are replaced in the form of a slit, the parts are modified by computer graphic technology or the like.

[0003]

The first montage photograph cannot be stored in many places to be searched, and the impression of the first discoverer with respect to the culprit depends on the manner in which the searcher is guided. It has the disadvantage of fading. In addition, the second montage photograph cannot be adopted except in a search place where there is a search person who can write a portrait. An investigator having the above-mentioned special skills is difficult to assign to all the places to be searched. There is a disadvantage that the impression fades.

In the third montage photograph, a large number of photographs can be classified and stored in a storage device, and each part can be easily replaced. But,
Since the above-mentioned montage photograph uses many image processing techniques for correction in image processing, the computer becomes large and it was impossible to reduce the size so that it could be brought to an investigation site. The place to be searched having the computer capable of performing the image processing is a limited place away from the crime scene, and has a disadvantage that the impression of the first discoverer of the criminal is weakened. Also, in the conventional montage photograph, even if the image is modified after replacing each part, if the impression is different, there is no other method than replacing the different parts. And while repeating this,
The first discoverer of the culprit has the disadvantage that the impression is diminished.

[0005] In order to solve the above problems, the present invention comprises a laptop personal computer that can be carried by an investigator to an investigation site and can be easily operated. It is an object to provide an image processing method and an image processing apparatus that can be manufactured. The present invention provides an image processing method and an image processing apparatus capable of obtaining an image of a face used in forensics by obtaining an image between two selected images having features of a face or a part. The purpose is to provide.

[0006]

According to the first aspect of the present invention, there is provided an image processing method comprising the steps of: providing a contour of a face drawn in a frame and control points provided in a mesh shape for each part constituting the face; A large number of face images composed of position information and names indicating the position of each part are classified and stored in the data library, and the classified face images corresponding to the images are stored from the data library. Searching and displaying multiple face images, selecting one face image that best matches the image from the multiple face images as the basic face, and then selecting one face characteristic for the part Is selected and displayed together with the basic face as a reference face, and the morphing process is performed on the basic face and the reference face to obtain an image of a face existing between any of these. Face image Characterized in that the Seisuru.

(Second Invention) An image processing method according to the present invention is characterized in that a contour of a face drawn in a frame and control points provided in a mesh shape for each part constituting the face and a position of each part are determined. A large number of face images composed of the position information and the names shown are classified and stored in a data library, and a plurality of face images are searched from the data library by searching for a classified face image matching the image. The image of the face is displayed, and the image of one face that best matches the image is selected as the basic face from the images of the plurality of faces, and then the image of one face characteristic of the part is selected. , A morphing process is performed on the basic face and the reference face to create an image of a face existing between any of the basic face and the reference face, and the face after the morphing process is performed. In the image of Characterized by making an image of the face of the imagewise by replacing each part.

(Third invention) An image processing method according to the present invention is characterized in that the outline of a face drawn in a frame and the control points provided in a mesh shape for each part constituting the face and the position of each part are determined. A large number of face images composed of the position information and the names shown are classified and stored in a data library, and a plurality of face images are searched from the data library by searching for a classified face image matching the image. The image of the face is displayed, and the image of one face that best matches the image is selected as the basic face from the images of the plurality of faces, and then the image of one face characteristic of the part is selected. , A morphing process is performed on the basic face and the reference face to create an image of a face existing between any of the basic face and the reference face, and the face after the morphing process is performed. In the image of Swapping each part, characterized in that to produce the image of the face of the imagewise by changing the image between the arbitrarily subjected to morphing processing part from the characteristics of the extremes for the part that is replaced each.

(Fourth Invention) The image processing method of the present invention provides a morphing process for a basic face and a reference face, a process for replacing each part, or an arbitrary change in the order of the morphing process for each part, thereby obtaining a face as an image. Is characterized by producing an image of

(Fifth invention) The image processing method of the present invention is characterized in that a face image obtained by the first to third inventions is modified by computer graphic processing into an image close to a painting. I do.

(Sixth invention) An image processing apparatus according to the present invention includes a control unit provided in a mesh shape for a contour of a face drawn in a frame and each part constituting the face, and a position of each part. A data library 112 that classifies and stores a large number of face images composed of the indicated position information and names, and selects a plurality of similar face images from the data library 112 by selecting an image. At the same time, a face selecting means 111 for selecting a basic face and a reference face characteristic of parts corresponding to the image, a morphing processing means 117 for creating an image of a face existing between any of the basic face and the reference face, Display means 115 for displaying the image of the face obtained by the face selecting means 111 and the morphing processing means 117. Characterized in that it is.

(Seventh invention) An image processing apparatus according to the present invention is capable of determining the outline of a face drawn in a frame and the control points provided in a mesh shape for each part constituting the face, and the position of each part. A data library 112 that classifies and stores a large number of face images composed of the indicated position information and names, and selects a plurality of similar face images from the data library 112 by selecting an image. At the same time, a face selecting means 111 for selecting a basic face and a reference face characteristic of parts corresponding to the image, a morphing processing means 117 for creating an image of a face existing between any of the basic face and the reference face, Morphing processing means 11
And a display unit 115 for displaying the face image obtained by the face selection unit 111 and the morphing processing unit 117. , An image of the face as the image is produced.

(Eighth Invention) An image processing apparatus according to the present invention is capable of determining the outline of a face drawn in a frame and control points provided in a mesh shape for each part constituting the face, and the position of each part. A data library 112 that classifies and stores a large number of face images composed of the indicated position information and names, and selects a plurality of similar face images from the data library 112 by selecting an image. At the same time, a face selecting means 111 for selecting a basic face suitable for the image and a reference face characteristic of the parts, and a part selecting means 1 for selecting and replacing any part of the basic face
And a morphing processing unit 117 that creates an image of a face existing between any of the basic face and the reference face, and arbitrarily deforms the part according to the characteristics of the parts.
And a display means 115 for displaying a face image obtained by the face selecting means 111, the parts selecting means 118, and the morphing processing means 117, and a face image as an image is produced. And

(Ninth Invention) In the image processing apparatus of the present invention, the number of control points is the same for a face existing between an arbitrary one of a basic face and a reference face and for a part arbitrarily deformed by the feature of the part. In addition, after the maximum coordinate position and the minimum coordinate position of the control point corresponding to the same position are prorated, correction processing is performed.

(Tenth invention) The parts in the image processing apparatus according to the present invention are characterized in that they include objects to be worn other than the facial features.

(Eleventh Invention) The face image produced by the image processing apparatus of the present invention is a montage image used for forensic investigation.

(Twelfth Invention) An image processing apparatus according to the present invention
It is a laptop personal computer.

[0018]

BEST MODE FOR CARRYING OUT THE INVENTION (First Invention) A data library of an image processing apparatus stores a large number of face images which are contained in one frame and in which control points are provided in a mesh shape. ing. For example, by moving one of the mesh-like control points, the control points in the vicinity also move, and the point information is changed to surface information to correct the mutual positions. Parts constituting the face or parts such as objects to be worn are composed of names and their position information, and can be replaced with substantially the same position on different faces. Further, the data library stores a large number of face images by classifying them as a frame together with the outline of the face and each part constituting the face. For example, the first discoverer of the criminal searches the data library for a face image classified according to the contour and image of the face, and causes the display device to display a plurality of face images. The first discoverer of the culprit selects, as a basic face, an image of one face that best matches the contour or image from the plurality of images of the face displayed on the display device.

Next, the first discoverer of the criminal selects an image of one face characteristic of the part from the plurality of images of the face or an image extracted by another classification. A reference face is displayed in parallel with the basic face. The classification of the image can be made arbitrarily based on the outline of the face, in the order of small face, large eye, etc. The basic face and the reference face are subjected to a morphing process used as a technique such as computer graphics, so that an image of a face existing between any of these is obtained. According to the first invention, a face image from the basic face or a face image from the reference face can be set between any of the basic face and the reference face having different points of view. For example, the image of the face from the 100% basic face is the basic face itself, and the image of the face from the 0% basic face is the reference face. When 50% is specified, an image intermediate between the basic face and the reference face is obtained. The first invention is, as described above, an alternative between the basic face and the reference face, and an image of the face between any of these can be obtained. Montage photos can be made.

(Second Invention) The second invention is different from the first invention in that a part of the face or a part which is a belonging can be replaced. For example, the first discoverer of the culprit may obtain an image as impressed by the image of the face that has been subjected to the morphing process with the basic face and the reference face, but if it is slightly different, by replacing each part, An image of the face closer to the first impression can be obtained.
The above parts are replaced with parts of the same name of the basic face based on the name and position information given to the parts. The image of the basic face is subjected to a process of transmitting the part based on the position information of the part, and then is replaced with a new part. According to the second aspect of the present invention, an image between the two faces can be obtained arbitrarily from the images of the two faces, and each part can be freely replaced. Therefore, for example, an image close to the impression of the first discoverer of the criminal can be obtained.

(Third invention) In the third invention, after replacing a part of the face or a part which is a belonging, a morphing process is performed so as to fall between any two extremes which are the characteristic of the part and change. This is different from the first and second aspects of the present invention. For example, the hair can be obtained by a morphing process between any of a case where the hair is the largest and a case where the hair is the least. Therefore, it can be said that the amount of hair is reduced to half of the case where the hair is the most, or to 20%. The third invention is each feature of the face other than the hair,
The same can be applied to the color of the complexion, the color of the skin, or the belongings of the criminal, so that an image closer to the first impression can be obtained. The image processing method according to the first to third aspects of the present invention uses a method of alternately selecting an image, a known technique of morphing, and replacement of parts to reduce the storage capacity of the data library. No facial images can be produced. In addition, since it is possible to alternate between two face images, or to replace parts in the selected face image, only by pressing the execution key or clicking the mouse, it is specialized in producing montage photos. A house is not required, and even the first discoverer of the perpetrator can create a montage photograph.

(Fourth Invention) According to the fourth invention, the order of the morphing process for the basic face and the reference face, the process of exchanging parts, or the morphing process for parts can be arbitrarily changed in the second and third inventions. . For example, the morphing process can be performed after replacing any part of the basic face and the reference face. Further, after the parts of the basic face and the reference face are exchanged, the morphing processing on the basic face and the reference face after the morphing processing on the parts can be performed again. As described above, by arbitrarily changing the order of the above processes, an image closest to the face left as an image can be easily and accurately created.

(Fifth Invention) The face image input to the data library of the present invention is basically the same as the photograph after the image processing because the photograph is taken in by the image scanner. However, as human psychology,
When the criminal is overhanged as a montage photograph, the general public tends to search for the person shown in the montage photograph. Therefore, in a fourth invention, the image of the face obtained by the first invention to the third invention is modified by computer graphic technology to make it an image close to a painting. For example, the images obtained by the first to third inventions are subjected to a process for changing the size of each pixel, a blur process, a process for changing the hit condition of light rays, a process for inserting a background, and the like. . According to the fourth invention, when a montage photograph formed by the above-described processing is viewed, the image is close to a picture, so that the image of the criminal remains in the mind and the arrest rate can be improved.

(Sixth Invention) An image processing apparatus according to a sixth invention is characterized in that:
A data library in which a large number of faces are classified and stored, a face selecting means capable of selecting an image of the displayed face, and a morphing processing means capable of creating an image existing between any two faces, Display means for displaying a face image. The data library is
A large number of face images composed of mesh contours and control points provided for each part constituting the face and position information indicating the name and position of each part are drawn in the frame. Classified and stored. For example, the position information of the mouth is a predetermined position in the face and includes information surrounding the mouth, so that the shape, thickness, and the like of the mouth can be changed.

The investigator selects a plurality of similar face images from the data library by operating the face selecting means in the computer while listening to the image of the first discoverer of the criminal, for example. I do. The investigator, with the help of the first discoverer of the culprit,
A face with a characteristic contour or image that gives the impression closest to the criminal is selected as a basic face, and an image of a face with characteristic parts is selected as a reference face. The morphing processing means creates an image of a face existing between any of the basic face and the reference face. For example, when the investigator specifies 50% in the morphing processing means, the contour of the face becomes the above-mentioned basic face, and the image of the face in which the other parts exist exactly in between the two is displayed on the display device. Thus, the fifth
According to the present invention, since an image of a face similar to both the basic face and the reference face can be instantly created, a high-quality montage photograph can be obtained before the impression of the first discoverer of the culprit fades. .

(Seventh invention) In the seventh invention, a part selecting means is added to the sixth invention. For example, the investigator creates an image of a face existing between any of the basic face and the reference face, and then selects a part by the part selecting means. Since the parts are stored in the data library with their names and position information attached, after being designated, the parts displayed on the display means are processed to be transparent, Other specified parts are replaced. According to the seventh aspect, since a part of the face or belongings can be easily replaced, an accurate montage photograph is produced.

(Eighth invention) In the eighth invention, the parts selection means is added to the sixth invention, and the morphing processing means can perform a process of turning parts into arbitrary parts. I have. For example, the mouth can be obtained by a morphing process between a small case and a large case. Also, the mouth can be anywhere between the thick and thin lips, or between any of the top and bottom.

(Ninth Invention) A face existing between any of the basic face and the reference face and a part arbitrarily deformed by the feature of the part are stored in the data library at the time when they are stored.
All have the same number of control points. Further, when creating the basic face and the reference face, the maximum coordinate position and the minimum coordinate position of the control point corresponding to the same position are equally divided. Thereafter, the face image created during any time is subjected to a correction process generally used in computer graphic technology. According to the ninth aspect, an arbitrary image between two faces or parts thereof can be obtained by a simple operation.

(Tenth Invention) The parts according to the seventh to ninth inventions include things to be worn other than the facial features.
According to the tenth aspect, the type, color, and shape of clothing worn, eyeglasses worn on the face, earrings, skin color, and the like can be changed.

(Eleventh Invention) An eleventh invention is a montage image used for a forensic investigation, and is useful when searching for identities such as a criminal, a runaway person, or an unknown body.

(Twelfth Invention) According to the twelfth invention, since a face having a desired impression is produced from the two faces in an alternative manner, the number of images of the data library to be input in advance can be reduced. A laptop-type personal computer that could be brought to the criminal investigation site could be obtained. In addition, the twelfth invention can be easily carried to an investigation site and can be operated by a non-expert, so that it exerts its power in an initial investigation.

[0032]

FIG. 1 is a block diagram for explaining an image processing apparatus according to a first embodiment of the present invention. In FIG. 1, the image processing apparatus includes, for example, a face selecting unit 111 for selecting a face on which a montage photograph is to be created,
A data library 112 storing a large number of face images each having the classified features;
Buffer memory 113 serving as a work area for temporarily storing image data of the face selected by the user and performing predetermined processing, and display control means for controlling display of the image data stored in the buffer memory 113 114
A display device 115 for displaying a face image or the like based on the control of the display control means 114;
Morphing setting means 116 for setting an arbitrary interval between two faces selected by the user, and morphing processing means 117 for performing processing for setting a face between two faces based on the morphing setting means 116
A part selecting means 118 for selecting a part of the face or an object to be worn; a parts fine adjusting means 119 for causing the morphing processing means 117 to process the part selected by the parts selecting means 118; A printing device 120 for printing the processed data stored in the buffer memory 113;
An image scanner 121 for reading a face image into the camera 12
And control means 122 for controlling each means.

In the present embodiment, by selecting two faces, creating a face between any two faces, exchanging parts, and making it between any two extremes of the parts, an unlimited number of pictures can be obtained. Since a large number of photographs can be produced, the storage capacity of the data library 112 can be significantly reduced. In the present embodiment, for example, an investigator and a first culprit of the criminal cooperate to produce the image while viewing the display device of the image processing apparatus. Hereinafter, in the present specification, the two persons are simply referred to as operators. The operator selects a face image stored in the data library 112 by the face selecting means 111. The operator selects two of the outline of the culprit's face, a basic face close to the image, and a reference face that is characteristic of the parts.

The data library 112 stores the names of the outlines of the face and the parts constituting the face,
This is a storage device that classifies and stores a large number of face images composed of position information indicating these positions. The outline of the face and the names of the parts constituting the face can be stored in the data library 112 so as to correspond to the characters. The operator inputs, for example, a character “round face” to display a plurality of round face images on the display device. At this time, data relating to a plurality of round face images is transferred from the data library 112 to the buffer memory 113. The images of the plurality of round faces stored in the buffer memory 113 are displayed on the display device 115 under the control of the display control means 114.

The operator looks at the display device 115 and selects one basic face that matches the contour and image of the face. Next, the operator searches for a face characteristic of the part. Also in this case, by inputting characters, images of a plurality of faces characteristic of the parts are displayed on the display device 115. The operator looks at the display device 115 and selects an image of a face characteristic of a part as one reference face. The selected basic face and reference face are simultaneously displayed on the display device 115. The operator compares the basic face and the reference face and determines which side is similar. Which side is similar to the above can be set from 100% to 0% by the morphing setting means 116. For example, 100% of the basic face is the basic face itself, and is a reference face itself rather than the 0% reference face. Further, the morphing setting means 116 continuously determines which side of the basic face and the reference face are similar, or determines whether the face is similar to the reference face by 10%.
Stepwise settings such as% can also be made.

The morphing processing means 117 is often used, for example, in computer graphic technology, and is a program for producing an intermediate product between two pictures. The present applicant has paid attention to the fact that the morphing processing means 117 can directly use a program used in computer graphics. That is, an image between any of the basic face and the reference face is created by the well-known morphing processing unit 117. The basic face created in this way is printed as it is as a montage photograph by the printing device 120.

When the operator needs to modify the basic face, the operator selects the parts of the face and the parts of his / her belongings with the parts selecting means 118. The parts selection means 118
As with the face selecting means 111, the selection can be made by inputting characters. In the part selection by the part selecting means 118, an image having the characteristics of the specified part is selected from the face images in the data library 112 and displayed on the display device 115 as a reference face.
The reference face having features in the parts is for replacing parts of the basic face. For example, when a nose is selected as a part, processing is performed to make the nose of the reference face the nose of the basic face. That is, based on the selected nose position information, the basic face is subjected to processing for transmitting the nose part to be replaced. Then, on the part of the nose that was transmitted,
The nose selected from the reference face is replaced.

Further, the operator can use the morphing processing means 117 to raise or lower the nose. The parts fine adjustment means 119 causes the morphing processing means 117 to process an object having an arbitrary height between a high nose and a low nose, for example. The data library 112 can input an image serving as a basic face or a reference face by the image scanner 121 as needed. Further, the control unit 122 controls a program for executing the processing of each unit.

FIG. 2 is a view for explaining the parts fine adjustment means used in the embodiment of the present invention. In FIG.
The part fine adjustment means is, for example, the hair / forehead adjustment means 211.
Eyebrow adjusting means 212, eye adjusting means 213, nose adjusting means 214, mouth adjusting means 215, cheek adjusting means 216
, A beard adjusting unit 217, a clothing adjusting unit 218, and an option 219. Option 219 is
Data library 1 for shapes of belongings such as glasses and ornaments
12 so that an intermediate product of the size and shape can be arbitrarily manufactured. Each adjusting means includes, for example, an amount adjusting means, a width adjusting means, a thickness adjusting means, a shape adjusting means, a position adjusting means, a length adjusting means, a size adjusting means, a thickness adjusting means, and the like. It can be adjusted continuously or stepwise. For example, the mouth adjusting means 215 can adjust continuously or stepwise from small to large, thick to thin, upper to lower, and so on. it can.

FIG. 3 is a flow chart for explaining an image processing apparatus according to an embodiment of the present invention. FIG. 4 shows FIG.
6A and 6B are flowcharts for explaining the image processing apparatus to which FIGS. FIG. 5 is a diagram for explaining an example of storing names assigned to one face, determined control points, and other information. FIG. 6 is a diagram for explaining an example of gender classification in one embodiment of the present invention. FIG. 7 is a diagram for explaining an example in which a large classification is subdivided according to an embodiment of the present invention. FIG. 8 is a diagram illustrating an example of a correspondence relationship between a character input and a data library identifier according to an embodiment of the present invention. FIG. 9A is a diagram for explaining an example of a display screen when first selecting a basic face or a reference face according to an embodiment of the present invention, and FIG. 9B is a diagram for transforming each part into an arbitrary shape. FIG. 4 is a diagram for explaining an example for performing the operation. FIG. 10 is a diagram illustrating an example of a state in which a basic face and a reference face are displayed on a display screen according to an embodiment of the present invention. FIG. 11 is a view for explaining an example of a display screen when replacing parts in one embodiment of the present invention. FIG. 12 is a view for explaining an example of fine adjustment of parts according to an embodiment of the present invention.
FIG. 13A shows an example of an image of a face intermediate between a basic face and a reference face, FIG. 13B shows an example of a face image in which a descending eye is selected when eyes are exchanged, and FIG. (C) is an example of an image of a face with a rising eye selected.

The flowcharts of FIGS. 3 and 4 will be described with reference to FIGS. 1 and 5 to 13. First, a program for creating a montage photograph is started. When the program is started, a menu is displayed on the display device 115 (see FIG. 1). The operator operates the display device 1
Looking at the above menu at 15, for example, click on "Make Montage" with the mouse. As shown in FIG. 9, the outline of the face or the features of the image are input as characters (step 311 (see FIG. 3)). The relationship between the character and the identifier is based on, for example, a table whose example is shown in FIGS. In such a relationship, by inputting characters, images and parts of the associated face are displayed.

By inputting the face outline or image feature, the display screen of the display device 115 becomes as shown in FIG. 9, and a plurality of male face images are displayed. The operator checks whether or not there is an image of the face having the input feature on the display screen (step 312). If the face image displayed on the display screen is different from the impression of the operator, the process returns to the previous step 311 in order to change and input the outline and the expression of the image. When the operator determines that a face image based on the input features is present on the display screen, the operator selects the face image closest to the impression as the basic face.
This selection by the operator is performed by face selection means 11 including a mouse.
1 (see FIG. 1) stores the data consisting of the image of the selected face in the data library 112 in the buffer memory 11.
3 is performed. That is, the face selecting means 111 selects a basic face from the data library 112 and displays it on the display device 115 (step 313).

In the above display screen, the operator can display the features related to parts such as eyes, eyebrows, and nose in FIG.
Input with characters as shown in (a) (step 31)
4). The selection of the basic face and the reference face and the selection of the parts can be performed while looking at the image of the face displayed on the display device 115 other than by inputting the characters. The operator checks whether there is a face of a part close to the impression of the face displayed on the display screen (step 315). If the operator has different impressions such as eyes, eyebrows, and nose, the operator returns to step 314 to input different characteristics or search the display device 115.

When the operator determines that a face image based on the input feature of the part is present on the display screen,
The face image closest to the impression is selected as the reference face. This selection by the operator causes the face selecting means 111 (see FIG. 1) to perform processing for transferring the data of the reference face selected in the data library 112 to the buffer memory 113. That is, as shown in FIG. 10, the display device 115 becomes a screen on which the basic face and the reference face are displayed (step 316). The operator looks at the screen shown in FIG. 10 and considers which of the basic face and the reference face is similar.
Next, the operator determines whether or not to adopt a position between the basic face and the reference face (step 317).

If the operator determines that the image is a face image between the basic face and the reference face, the operator selects the degree of similarity, for example, a number in FIG. 10 (step 31).
8). This number is, for example, if you choose 90%
Means 90% similar to the basic face and 10% similar to the reference face. If 10% is selected, it means similar to the basic face by about 10% and similar to the reference face by 90%. doing. After being selected by the face selecting unit 111, the morphing processing unit 117 performs a process of creating a face biased by the selected number using a well-known computer graphic technique (step 319).

The image shown in FIG. 13A is based on the outline of the basic face in FIG.
That is, an image intermediate between the basic face and the reference face is shown. The operator determines whether to temporarily fix the morphing-processed image (Step 3).
20). If the operator decides not to temporarily fix,
Returning to step 318, the distribution of the basic face and the reference face is changed. If it is close to 100% of the basic face, it is the basic face itself, and 0%, which does not resemble the basic face at all, becomes the reference face (step 321).

The operator finally temporarily fixes the basic face on the display screen (step 322). Further, the operator examines whether or not to replace parts such as eyes, nose and mouth (step 323). If the operator determines that the parts should be replaced, the parts are designated by the parts selecting means 118 (see FIG. 1) (step 324). If a face that matches the impression cannot be obtained simply by replacing parts, the operator finely adjusts the parts (step 325). For example, in the case of hair, the fine adjustment of the parts can be arbitrarily made between a large amount and a small amount. Any amount between high and low hair is achieved by morph fig processing techniques (step 326).

The operator determines whether or not to replace parts other than the hair (step 327). The operator repeats the process as long as there is a part to be replaced (steps 328 to 330). The image shown in FIG. 13 (b) is an example in which the eyes are exchanged as parts and then the morphing process is performed to lower the eyes. Further, the image shown in FIG. 13C is an example in which the morphing process is performed in a rising direction, contrary to the above. When parts are not replaced or finely adjusted, the operator determines a face (step 331). For the replacement of parts, for example, when the part is clicked with the mouse, the part part shown in FIG. 9B is displayed, and the mark O can be clicked with the mouse. The operator determines whether to print the face image as a photograph or to make a decoration like a picture (step 332).

When the operator wants to process a face image as a picture, the operator performs a modification process (not shown) (step 333). The modification process is a well-known means in the computer graphic arts, and uses this well-known means. In addition, the above-mentioned modification processing is sold as various programs, such as one that changes the size of a pixel, one that blurs a pixel, one that changes the way of applying light, and one that inserts a background. After performing the decoration processing on the face, processing for fixing the face is performed (step 334). The fixed face is
The printing is performed by the printing device 120 (see FIG. 1) in the same manner as when the decoration process is not performed (step 335).

FIG. 11 shows a screen for replacing parts and finely adjusting the parts. FIG. 11 shows a case where the hair and the forehead are exchanged, and it can be seen that the hair and the forehead are designated by the thick lines.
In addition, items for finely adjusting the parts are displayed by selecting the hair and forehead. In the case of hair, whether the hair is large or small, the amount can be set to an arbitrary amount by moving a cursor with a mouse. In the case of the forehead, whether it is wide or narrow, the cursor can be similarly moved to a desired width by moving the mouse. FIG. 12 shows an example of an item for finely adjusting a part. In the case of an eyebrow, the item is thick to thin, the character is V to V, and any item between the upper and lower positions. be able to.

Next, another method for obtaining an arbitrary image between faces or parts will be described. The images of the face and the parts stored in the data library are composed of the same number of control points, and the maximum size and the minimum size are represented by coordinate points. For example, focusing on the left side of the mouth, the largest coordinate point is read as the first coordinate point, and the smallest coordinate point is read as the second coordinate point. Then, of the read first and second coordinate point data, an operation is performed according to a desired ratio between the first to n-th identical coordinate point data. The above calculation is based on the following equation. (1) X _C1 = ｛R · X _Ai + (10
0−R) · X _Bi ｝ / 100 (2) Y _C1 = {R · X _Ai + (100−R) · Y _Bi } / 1
00 However, to not coordinate points A ₁ coordinate point data A _n X _A1,
X _A2 , X _A3 ,... X _An , and coordinate point data of the coordinate points B ₁ to B _n are X _B1 , X _B2 , X _{B3 ,.} Coordinate point data of points C ₁ to C _n (X _C1 , Y _C1 ), (X _C2 , Y _C2 ),.
-, (X _Cn , Y _Cn ) R is 0% to 100% in rate The coordinate point data of each control point calculated by these equations is displayed on the display device 115 as a mouth having a desired rate.

As described above, the present embodiment has been described in detail.
It is not limited to the above embodiment. Various design changes can be made without departing from the present invention described in the appended claims. The flowchart in the present embodiment is merely an example, and a quick and accurate face image can be obtained by arbitrarily changing the order of the morphing process, the replacement process of the parts, and the morphing process on the parts for the basic face and the reference face. Can be. In addition, the fact that an accurate face image can be obtained quickly means that a face image that matches the image can be obtained before the image based on the first impression disappears. Although each block means of the present invention is not described in detail in the embodiments, it can be constituted by a known or known means. The morphing processing means is a publicly known or well-known means based on computer graphics or the like, and is not described in detail in the embodiments.

[0053]

According to the present invention, an arbitrary face can be created between the basic face and the reference face, so that the capacity of the storage device is reduced, and the data processing is quick and the operation is simplified. This makes it easy to create montage photos without relying on specialists. Therefore, the image processing apparatus of the present invention
It is easy to carry and operate, and is suitable for producing a montage photograph at a crime scene. According to the present invention, the face and belongings can be easily replaced as parts,
Since various types of parts such as quantity, width, thickness, shape, position, length, size, thickness, etc. can be arbitrary, the capacity of the storage device can be reduced as described above, Processing can be accelerated. As a result, it can be applied to forensic investigations targeting various kinds of investigations, and because of its low cost, it can be arranged at any investigation place and can be operated without requiring training of the investigator.

[Brief description of the drawings]

FIG. 1 is a block diagram illustrating an image processing apparatus according to a first embodiment of the present invention.

FIG. 2 is a diagram for explaining a part fine adjustment unit used in the embodiment of the present invention.

FIG. 3 is a flowchart illustrating an image processing apparatus according to an embodiment of the present invention.

FIG. 4 is a flowchart for explaining the image processing apparatus to which FIGS. 3A and 3A are connected.

FIG. 5 is a diagram for explaining an example when a name given to one face, a determined control point, and other information are stored.

FIG. 6 is a diagram for explaining an example of gender classification in one embodiment of the present invention.

FIG. 7 is a diagram for explaining an example of subclassification from large classification in one embodiment of the present invention.

FIG. 8 is a diagram illustrating an example of a correspondence relationship between a character input and a data library identifier according to an embodiment of the present invention.

FIG. 9A is a view for explaining an example of a display screen when a basic face or a reference face is first selected according to an embodiment of the present invention, and FIG. It is a figure for explaining an example for transforming into.

FIG. 10 is a diagram illustrating an example of a state in which a basic face and a reference face are displayed on a display screen according to an embodiment of the present invention.

FIG. 11 is a diagram for explaining an example of a display screen when replacing parts in one embodiment of the present invention.

FIG. 12 is a diagram illustrating an example of fine adjustment of a part according to an embodiment of the present invention.

13A is an example of an image of a face intermediate between a basic face and a reference face according to an embodiment of the present invention, and FIG. 13B is an example of an image of a face in which a descending eye is selected when eyes are exchanged. , (C) is an example of a face image with a rising eye selected.

[Explanation of symbols]

 111 ... face selection means 112 ... data library 113 ... buffer memory 114 ... display control means 115 ... display device 116 ... morphing setting means 117 ... morphing processing means 118 ...・ Parts selection means 119 ・ ・ ・ Parts fine adjustment means 120 ・ ・ ・ Printing device 121 ・ ・ ・ Image scanner 122 ・ ・ ・ Control means

Claims (12)

[Claims] An outline of a face drawn in a frame and control points provided in a mesh shape for each part constituting the face, and a large number of position information and names indicating the position of each part. The images of the faces are classified and stored in a data library, a plurality of images of the face are displayed by searching the images of the face corresponding to the image from the data library, and the plurality of faces are displayed. Select the image of one face that best fits the image from among the images of the above as the basic face, and then select the image of one face that has features in the parts,
A face image is displayed as a reference face together with the basic face, and a morphing process is performed on the basic face and the reference face to obtain a face image existing between any of the basic face and the reference face. An image processing method comprising: 2. A contour comprising a face drawn in a frame and control points provided in a mesh shape for each part constituting the face, and a plurality of position information and names indicating the position of each part. The images of the faces are classified and stored in a data library, a plurality of images of the face are displayed by searching the images of the face corresponding to the image from the data library, and the plurality of faces are displayed. Select the image of one face that best fits the image from among the images of the above as the basic face, and then select the image of one face that has features in the parts,
It is displayed together with the basic face as a reference face, and by performing morphing processing on the basic face and the reference face, an image of a face existing between these arbitrary faces is created, and the face after the morphing processing is performed An image processing method characterized by producing an image of a face as an image by replacing each part in the image. 3. A plurality of contours of a face drawn in a frame and control points provided in a mesh shape for each part constituting the face, and positional information and a name indicating the position of each part. The images of the faces are classified and stored in a data library, a plurality of images of the face are displayed by searching the images of the face corresponding to the image from the data library, and the plurality of faces are displayed. Select the image of one face that best fits the image from among the images of the above as the basic face, and then select the image of one face that has features in the parts,
It is displayed together with the basic face as a reference face, and by performing morphing processing on the basic face and the reference face, an image of a face existing between these arbitrary faces is created, and the face after the morphing processing is performed An image characterized in that each part in the image is replaced, and a morphing process is performed on the part based on the extreme characteristics of each of the replaced parts to change the image to an image between any of the parts, thereby producing a face image as an image. Processing method. 4. The image processing method according to claim 2, wherein the order of the morphing process for the basic face and the reference face, the replacement process of each part, or the morphing process for each part is arbitrarily changed.
An image processing method characterized by producing an image of a face according to an image. 5. An image processing method, characterized in that a face image obtained by the image processing method according to claim 1 is modified by computer graphic processing to obtain an image close to a painting. 6. A contour comprising a face drawn in a frame and control points provided in a mesh shape for each part constituting the face, and a plurality of position information and names indicating positions of the parts. A data library in which face images are classified and stored, and a plurality of similar face images are selected from the above data library by selecting an image, and the characteristic faces and basic faces and parts that match the image are selected. A face selection means for selecting a certain reference face; a morphing processing means for creating an image of a face existing between any of the basic face and the reference face; and a morphing processing means for generating a face image obtained by the face selection means and the morphing processing means. An image processing apparatus comprising: display means for displaying an image; and producing an image of the face as the image. 7. A plurality of contours of a face drawn in a frame and control points provided in a mesh shape for each part constituting the face, and positional information and a name indicating the position of each part. A data library in which face images are classified and stored, and a plurality of similar face images are selected from the above data library by selecting an image, and the characteristic faces and basic faces and parts that match the image are selected. Face selecting means for selecting a certain reference face; morphing processing means for creating an image of a face existing between any of the basic face and the reference face; and arbitrary parts in the basic face obtained by the morphing processing means And a display unit for displaying a face image obtained by the face selection unit and the morphing processing unit. An image processing apparatus characterized by making an image of the face imagewise. 8. A large number of contours of a face drawn in a frame and control points provided in a mesh shape for each part constituting the face, and positional information and a name indicating the position of each part. A data library in which face images are classified and stored, and a plurality of similar face images are selected from the above data library by selecting an image, and the characteristic faces and basic faces and parts that match the image are selected. A face selecting means for selecting a certain reference face, a part selecting means for selecting and replacing an arbitrary part in the basic face, and creating an image of a face existing between any of the basic face and the reference face, Morphing processing means for arbitrarily deforming parts according to the characteristics of each of the parts, and the morphing processing means obtained by the face selecting means, parts selecting means, and morphing processing means Display means for displaying an image, includes a, an image processing apparatus characterized by making an image of the face imagewise. 9. A face existing between any of the basic face and the reference face and a part arbitrarily deformed by the feature of the part have the same number of control points and the maximum number of control points corresponding to the same position. 9. The image processing apparatus according to claim 6, wherein a correction process is performed after the coordinate position and the minimum coordinate position are divided. 10. The apparatus according to claim 6, wherein said parts include objects to be worn in addition to facial features.
The image processing apparatus according to any one of the preceding claims. 11. An image processing apparatus, wherein the face image created by the image processing apparatus according to claim 6 is a montage image used for a forensic investigation. 12. The image processing apparatus according to claim 6, wherein the image processing apparatus is a laptop personal computer.