JP4837602B2 - Image trimming apparatus and method, and program - Google Patents

Description

  The present invention relates to an image trimming apparatus and method for setting a trimming region including a face for an image including a face, and a program for causing a computer to execute the image trimming method.

  When applying for a passport or license, or creating a resume, it is often required to submit a photo of a predetermined output standard (hereinafter referred to as an ID photo) showing the person's face. For this reason, a photo room is provided for taking pictures of the user, and the ID photo sheet is created by photographing the user sitting on the chair in the photo room and recording the face photo image for the ID photo of the user on the sheet. ID photo automatic creation devices that automatically do this have been used. Since such an automatic creation device is large and has limited installation locations, it is inconvenient for a user to search for the location where the automatic creation device is installed in order to obtain an ID photo. It is.

For this reason, a method has been proposed in which a face, eyes, and the like are detected from an image, and a trimming region is set based on the detected face and eyes. For example, Patent Document 1 proposes a method of determining a rotation angle of a face included in an image in the image plane from the center position of both eyes and correcting the face rotation to set a trimming region. In Patent Document 2, face detection, eye detection, and pupil detection are performed from an image, the positions of both pupils and the distance between both pupils are acquired, and the positions of both pupils and the distance between both pupils are obtained. A method for setting a trimming region by using it has been proposed. Further, in Patent Document 3, there is a method in which a straight line connecting the center point between both eyes, the center point of the nose, and the center point of the mouth is set as a trimming center line, and an area that is symmetrical about the trimming center line is set as a trimming area. Proposed. According to these methods, a trimming region including a face can be set in an image with the center of the face as a reference.
JP 2003-319169 A JP 2005-242641 A JP 2006-139369 A

  By the way, the face included in the image is not necessarily a face facing front, but may be facing up, down, left and right. Although such a face facing up, down, left and right is not suitable as an ID photo, for example in school photos, it may be preferable to face slightly diagonally. There is a demand for trimming by setting a trimming area. For this reason, it is conceivable to set a trimming region in an image using the methods described in Patent Documents 1 to 3 above.

  Here, since the methods described in Patent Documents 1 to 3 use the center position of both eyes as a reference for setting the trimming region, the trimming region is set so that the face facing the front is centered. Can be set. However, since the center position of both eyes is not necessarily located at the center of the face when the face is faced up, down, left, and right, the center of the trimming region is the center of the face when the methods of Patent Documents 1 to 3 are used. As a result, the face is not positioned at the center of the trimming area.

  The present invention has been made in view of the above circumstances, and an object thereof is to allow a trimming region to be appropriately set regardless of the face orientation.

An image trimming apparatus according to the present invention is an image trimming apparatus that sets a trimming region including a face to an image including a face.
Face detection means for detecting a face included in the image;
Feature quantity acquisition means for acquiring the feature quantity of the face from the detected face;
Center position setting means for setting a center position of the face based on the detection result of the face and the feature amount;
Trimming area setting means for setting a trimming area for the image with the center position as the center is provided.

The image trimming apparatus according to the present invention further includes orientation detection means for detecting the orientation of the face based on the feature amount.
The center position setting means may be means for setting the center position of the face based on the orientation of the face.

Further, in the image trimming apparatus according to the present invention, the feature amount acquisition means is a means for acquiring the center position of both eyes included in the face, the position of the nose and the left and right end positions of the face as the feature amount,
The center position setting means is a means for setting the center position of the face on the basis of any one of the center positions of the eyes, the position of the nose and the left and right ends of the face, or the center of gravity of these positions. Also good.

An image trimming method according to the present invention is an image trimming method for setting a trimming region including a face to an image including a face.
Detecting a face contained in the image,
Obtaining a feature amount of the face from the detected face;
Based on the detection result of the face and the feature amount, the center position of the face is set,
A trimming region is set in the image around the center position.

  The image trimming method according to the present invention may be provided as a program for causing a computer to execute the image trimming method.

  According to the present invention, a face included in an image is detected, and a feature amount of the face is acquired from the detected face. Then, the center position of the face is detected based on the face detection result and the feature amount, and a trimming region is set around the center position. For this reason, the center position of the face can be appropriately detected regardless of the orientation of the face, and as a result, the trimming region can be set so that the face is located at the center. This makes it possible to trim the face used as the ID photo from the snap photo image without taking the ID photo.

  Hereinafter, embodiments of the present invention will be described with reference to the drawings. FIG. 1 is a schematic block diagram showing the configuration of an image trimming apparatus according to an embodiment of the present invention. As shown in FIG. 1, the image trimming apparatus according to the present embodiment includes an image input unit 1 that receives input of image data S0 representing an image including a face, and an image represented by the image data S0 (hereinafter, the image is also an image). The face detection unit 2 that detects a face from S0, the feature amount acquisition unit 3 that acquires a feature amount from the face detected by the face detection unit 2, and the face orientation are detected. An orientation detection unit 4, a center position setting unit 5 for setting the center position of the face in the image, a trimming region setting unit 6 for setting a trimming region around the center position of the face, and an image from the image S0 by the trimming region. A trimming unit 7 for trimming.

  The image input unit 1 includes a media drive that reads image data S0 recorded on the medium from the medium, and various interfaces that receive input of the image data S0 transmitted via the network. The image data S0 may be acquired by an imaging device such as a digital camera, or may be acquired by photoelectrically reading an image recorded on a film or a document.

  The face detection unit 2 uses a template matching method, a method using a face discriminator obtained by machine learning learning using a large number of sample images of a face, and a neural network described in Japanese Patent Laid-Open No. 5-282457. A face position on each image is detected as a face area by a technique or the like. When the face is not included in the image S0, the face detection unit 2 outputs that the face is not detected.

  The feature amount acquisition unit 3 acquires, as a feature amount, feature points of the face, which is a main face component constituting the face T0 included in the face area detected by the face detection unit 2. Specifically, for example, by using a template matching method using a template of each facial part or a method using a discriminator for each facial part obtained by machine learning learning using a large number of sample images of facial parts. The coordinates of the position of the eyes, the position of the left and right eyes, the position of the nose, the position of the top of the head, the position of the tip of the chin, and the position of the left and right edges of the face are acquired as feature amounts. Further, by calculating the positions of both eyes from the positions of the left and right eye heads and left and right eye corners, the coordinates of the positions of both eyes are also acquired as feature amounts. Specifically, the midpoint between the left and right eye heads and the corners of the eyes is calculated as the left and right eye positions.

  The direction detection unit 4 compares the positional relationship between the positions of both eyes and the position of the nose among the feature amounts acquired by the feature amount acquisition unit 3 with a predetermined standard positional relationship of the face facing the front, The difference in the positional relationship between the horizontal direction and the vertical direction of the image is calculated. Then, based on the difference in the positional relationship in the left-right direction, the left-right orientation of the face T0 is calculated as a value between −90 degrees and 90 degrees. Further, the vertical direction of the face T0 is calculated as a value between −90 degrees and 90 degrees based on the difference in the positional relationship in the vertical direction. The calculated left and right angles are D1 and the up and down angles are D2.

  The center position setting unit 5 calculates the center position C1 between the eyes, the nose position C2, and the center position C3 of the center position C3 of the left and right end positions of the face as reference positions for setting the center position of the face. Here, since the positions of both eyes are detected by the xy coordinates of the left and right eyes, if the coordinates of the right eye are (xr, yr) and the coordinates of the left eye are (xl, yl), the center position C1 between the eyes. The coordinates (x1, y1) are calculated as ((xr + xl) / 2, (yr + yl) / 2). The coordinates (x2, y2) of the nose position C2 are the coordinates of the detected nose position itself. Since the positions of the left and right edges of the face are detected by the xy coordinates of the left and right edges, if the coordinates of the right edge are (xer, yer) and the coordinates of the left eye are (xel, yel), the center C3 of the left and right edges The coordinates (x3, y3) are calculated as ((xer + xel) / 2, (yer + yel) / 2). The coordinates (x4, y4) of the center-of-gravity position C4 are calculated as ((x1 + x2 + x3) / 3, (y1 + y2 + y3) / 3).

  Furthermore, the center position setting unit 5 corrects the coordinates of the reference position, which is the center of gravity position C4, according to the face direction, and sets the face center C5. Specifically, coordinates (x5, y5) are calculated by the following equation (1), and the calculated coordinates (x5, y5) are set as the center position C5 of the face. In Expression (1), the face left-right radius R1 is ½ of the distance between the left and right edges of the face. Further, the face vertical radius R2 is ½ of the distance between the top of the head and the chin.

x5 = x4 + R1 × D1 / 90
y5 = y4 + R2 × D2 / 90 (1)
The trimming area setting unit 6 sets the trimming area in the image S0 so that the center of the trimming area matches the center position C5 of the face and the face to be trimmed fits in the trimming area.

  The trimming unit 7 cuts out a trimming area from the image S0.

  Next, processing performed in the present embodiment will be described. FIG. 2 is a flowchart showing processing performed in the present embodiment. Here, a process of trimming all the faces included in the image S0 will be described. When the image data S0 is read by the image input unit 1, the apparatus starts processing, and the face detection unit 2 detects a face from the image S0 (step ST1). Then, it is determined whether or not a face is included in the image S0 (step ST2), and if step ST2 is negative, the process is terminated.

  When step ST2 is affirmed, the face to be processed is set as the first face (step ST3), the feature amount acquisition unit 3 acquires the feature amount of the face to be processed (step ST4), and the direction detection unit 4 further. Detects the orientation of the face (step ST5).

  The center position setting unit 5 sets the center position C5 of the face as described above (step ST6), and the trimming region setting unit 6 sets a trimming region on the image S0 (step ST7). Then, it is determined whether trimming areas have been set for all faces (step ST8). If step ST8 is negative, the face to be processed is set as the next face (step ST9), and the process returns to step ST4. Then, the processing after step ST4 is repeated.

  When step ST8 is affirmed, the trimming unit 7 trims the trimming area (step ST10), and the process ends.

  FIG. 3 is a diagram for explaining the setting of the trimming area according to the present embodiment. When the trimming area is set with reference to the center of both eyes as in the methods described in Patent Documents 1 to 3, when the face is facing the front as shown in FIG. By positioning the center of the trimming area A1 at O, the trimming area A1 can be set so that the face is located at the center of the trimming area A1.

  However, when the face is inclined, if the trimming area is set with reference to the center position O of both eyes, the face is not positioned at the center of the trimming area A1 as shown in FIG.

  In the present embodiment, since the center position C5 of the face is detected according to the facial feature amount, particularly the face orientation, even if the face is facing sideways as shown in FIG. By positioning the center of the trimming area A1 at the center position C5, the trimming area A1 can be set so that the face is located at the center.

  In the above embodiment, the center position C5 of the face is set with the center C1 between the eyes, the position C2 of the nose and the center C3 of the center C3 of the position of the left and right ends of the face as the reference position. The center position C5 of the face may be set using any one of the center C1 between the eyes, the position C2 of the nose and the center C3 of the position of the left and right ends of the face without calculating the position C4 of the center of gravity.

  In the above embodiment, the coordinates of the positions of the right and left eyes, the positions of the left and right eyes, the position of the nose, the position of the top of the head, the position of the tip of the chin, and the position of the left and right edges of the face are acquired as feature amounts. Although the center position C5 of the face is set using the feature amount, the feature amount to be acquired is not limited to these. For example, the position of the left and right edges of the face, the position of the top of the head, and the position of the tip of the chin are obtained as feature quantities, the face size is estimated from these positions, and the center position of the region defined by the face size is The center position of the face may be set.

  Moreover, in the said embodiment, when the face detection part 2 detects a face, the position of both eyes may be detected. In such a case, the positions of both eyes detected by the face detection unit 2 may be acquired as feature amounts and used for setting the center position C5 of the face.

  In the above embodiment, the feature amount acquisition unit 3 is characterized by the coordinates of the positions of the left and right eyes, the positions of the left and right eyes, the position of the nose, the position of the top of the head, the position of the tip of the jaw, and the positions of the left and right edges of the face. Although it is acquired as a quantity, there are cases where all of these feature quantities cannot be acquired. In such a case, the feature quantity that the operator lacks may be input from a not-shown input unit by manual operation.

  In addition, since the position of both eyes can often be detected, when only the position of both eyes is acquired as a feature amount, the position of the nose, the position of the top of the head, the position of the chin, and the left and right sides of the face are statistically calculated from the positions of both eyes. The position of the end may be estimated. Further, the face size may be estimated from the positions of both eyes. If only the position of the top of the head is acquired as the feature amount, the position of the jaw may be statistically estimated from the position of the top of the head, and in this case, the position of the face is determined from the position of the top of the head and the position of the chin. The size can be estimated.

  Although the apparatus according to the first embodiment of the present invention has been described above, the computer includes the face detection unit 2, the feature amount acquisition unit 3, the orientation detection unit 4, the center position setting unit 5, and the trimming region setting unit 6 described above. A program that functions as means corresponding to the above and causes the processing shown in FIG. 2 to be performed is also one embodiment of the present invention. A computer-readable recording medium in which such a program is recorded is also one embodiment of the present invention.

1 is a schematic block diagram showing the configuration of an image trimming apparatus according to an embodiment of the present invention. A flowchart showing processing performed in the present embodiment The figure for demonstrating the setting of the trimming area | region of this embodiment

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 Image input part 2 Face detection part 3 Feature-value acquisition part 4 Direction detection part 5 Center position setting part 6 Trimming area setting part 7 Trimming part

Claims (4)

In an image trimming apparatus that sets a trimming region including a face to an image including a face,
Face detection means for detecting a face included in the image;
Feature amount acquisition means for acquiring, from the detected face, the position of both eyes included in the face, the position of the nose, and the positions of the left and right ends of the face as feature amounts;
Orientation detection means for detecting the orientation of the face based on the feature amount;
The center position between the eyes, the position of the nose and the positions of the left and right edges of the face, or the center position of the eyes, the position of the nose and the center of gravity of the positions of the left and right edges of the face Center position setting means for setting the center position of the face by correcting the reference position according to the orientation of the face as a reference position for setting the center position of the face;
An image trimming apparatus comprising trimming area setting means for setting a trimming area for the image with the center position as a center.   The orientation detection means compares the positional relationship between the positions of the eyes and the nose with a positional relationship of a predetermined standard frontal face, and detects the orientation of the face based on the comparison result. The image trimming apparatus according to claim 1, wherein the image trimming apparatus is a means. In an image trimming method for setting a trimming region including a face in an image including a face,
Detecting a face contained in the image,
From the detected face, the position of both eyes included in the face, the position of the nose and the positions of the left and right ends of the face are acquired as feature amounts ,
Detecting the orientation of the face based on the feature amount;
The center position between the eyes, the position of the nose and the positions of the left and right edges of the face, or the center position of the eyes, the position of the nose and the center of gravity of the positions of the left and right edges of the face A reference position for setting the center position of the face, correcting the reference position according to the orientation of the face, and setting the center position of the face,
An image trimming method, wherein a trimming region is set in the image with the center position as a center. In a program for causing a computer to execute an image trimming method for setting a trimming region including a face to an image including a face,
Detecting a face included in the image;
A procedure for acquiring, from the detected face, the position of both eyes included in the face, the position of the nose and the positions of the left and right ends of the face as feature amounts ;
Detecting the orientation of the face based on the feature amount;
The center position between the eyes, the position of the nose and the positions of the left and right edges of the face, or the center position of the eyes, the position of the nose and the center of gravity of the positions of the left and right edges of the face A reference position for setting the center position of the face, correcting the reference position according to the orientation of the face, and setting the center position of the face;
And a procedure for setting a trimming area in the image with the center position as a center.

Images