JP4327317B2 - Person image photographing method and person image photographing apparatus - Google Patents

Description

【００５２】
上記表１から明らかなように、黒髪の揚合はＢ輝度がブルーバックのＢ輝度とかけ離れているため、Ｂ輝度基準のみで頭部位置を精度よく検出できる。また、白髪、無頭髪の揚合、Ｂ輝度がブルーバックのＢ輝度とほとんど同じため、Ｂ輝度基準では検出できないが、Ｒ輝度がブルーバックのＲ輝度とかけ離れているため、Ｒ輝度基準により頭部位置を精度よく検出できることがわかる。
【００５３】
なお、前記実施の形態では、被撮影者の背景としてブルーバックを用い、ブルーバックのＢ輝度およびＲ輝度を基準として被撮影者の頭部位置を検出する場合について説明したが、本発明はこれに限定されるものでなく、たとえば、被撮影者の背景として赤色や緑色などの他の特定色を用いてもよい。その場合、被撮影者の背景である特定色に対して高輝度を示す色成分輝度および低輝度を示す色成分輝度を基準として被撮影者の頭部位置を検出する。
【００５４】
【発明の効果】
以上詳述したように本発明によれば、被撮影者が黒髪でない場合でも、被撮影者の頭部位置を精度よく検出することができ、被撮影者の顔の位置を撮影画像内で理想状態になるように配置できる人物画像撮影方法および人物画像撮影装置を提供できる。
【図面の簡単な説明】
【図１】本発明の実施の形態に係る人物画像撮影装置の外観構成を概略的に示す構成図。
【図２】図１に示した人物画像撮影装置の全体的な構成を示すブロック図。
【図３】被撮影者の頭頂位置を検出する際に用いる頭部の探索法を示した概念図。
【図４】処理動作を説明するためのフローチャート。
【図５】処理動作を説明するためのフローチャート。
【符号の説明】
１……被撮影者
２……照明器
３……ビデオカメラ（撮影手段）
４……撮影者（操作者）
５……表示器（モニタディスプレイ）
６……制御台（制御部）
１１……ビデオキャプチャ部
１２……ビデオアクセラレート部
１３……ＣＰＵ部
１５……大容量記憶媒体
１６……記憶媒体
１７……リーダライタ部
２０……画像印刷装置（出力手段）[0001]
BACKGROUND OF THE INVENTION
The present invention relates to a person image photographing method and a person image photographing apparatus for photographing an upper body image including at least a face of a person to be photographed, for example, for the purpose of creating an identification photograph to be affixed to an identification card or various licenses. About.
[0002]
[Prior art]
In general, in this type of human image photographing device, it is necessary to consider that the size and position of the face of the person to be photographed have a certain ratio (ideal state) in the photographed image. Conventionally, for example, blue (hereinafter referred to as a blue background) is used as a background of a subject, and a large area is photographed in advance in view of variations in the height of the subject, and the blue component luminance of the blue background (hereinafter, referred to as a blue background). The head position is detected from the top position and the left and right head positions of the subject on the basis of the brightness (referred to as B brightness), and the face position of the subject is in an ideal state in the photographed image based on the detected head position. A method of cutting out the upper body image of the subject from the photographed image was used.
[0003]
[Problems to be solved by the invention]
Assuming that the person to be photographed is a general Japanese with black hair, the head position of the person to be photographed (black hair part) is lower than the B brightness of the blue background, so the head position is detected with high accuracy. Is possible.
[0004]
However, due to the recent increase in the population of foreigners and aging in Japan, the possibility that the subject is not black hair is also fully considered. For example, when the subject is white hair or headless hair, the B brightness at the head position becomes high, and therefore the head position cannot be detected with high accuracy based on the blue back B brightness.
[0005]
Therefore, the present invention can accurately detect the position of the head of the photographed person even when the photographed person is not black hair, and is arranged so that the position of the face of the photographed person is in an ideal state in the photographed image. An object of the present invention is to provide a person image photographing method and a person image photographing device capable of performing the same.
[0006]
[Means for Solving the Problems]
The human image capturing method of the present invention includes a first step of capturing an image including at least a face of a subject having a specific color as a background, and an image captured by the first step. The second step for obtaining the color component luminance indicating high luminance and the color component luminance indicating low luminance for the specific color as the background, and the color component luminance obtained in the second step by the binary search method. a third step of detecting the head position of the person to be photographed as a reference, from the third image captured by the first step based on the head position of the person to be photographed, which is detected by the step wherein a fourth step of cutting out an image of a predetermined region of the person to be photographed, and a fifth step of outputting an image of a predetermined region of the person to be photographed cut out by the fourth step as a captured image .
[0007]
In the human image photographing method of the present invention, the specific color of the background is blue, the color component luminance indicating high luminance with respect to the specific color is blue component luminance, and the low luminance with respect to the specific color. The color component luminance shown is a red component luminance.
[0008]
Further, the human image photographing device of the present invention includes a photographing means for photographing an image including at least the face of the photographed person with a specific color as a background, and an image photographed by the photographing means in the background of the photographed person. Luminance calculation means for obtaining a color component brightness indicating a high brightness and a color component brightness indicating a low brightness for a specific color, and by means of a binary search method, the color component brightness determined by the brightness calculation means is used as a reference. Head position detection means for detecting the head position of the photographer, and the subject from the image photographed by the photographing means based on the head position of the subject photographed by the head position detection means An image cutting means for cutting out an image of the predetermined area, and an output means for outputting the image of the predetermined area of the photographed person cut out by the image cutting means as a taken image.
[0009]
Further, in the human image photographing device of the present invention, the specific color of the background is blue, the color component luminance indicating high luminance with respect to the specific color is blue component luminance, and the low luminance with respect to the specific color. The color component luminance shown is a red component luminance.
[0010]
According to the present invention, for example, blue (blue back) is used as the background of the subject, the head position of the subject is detected based on the blue component luminance and the red component luminance of blue, and the detected subject is detected. Based on the photographer's head position, by cutting out the upper body image of the photographed person so that the face position of the photographed person is in an ideal state, even if the photographed person is not black hair, The position of the part can be detected with high accuracy, and the position of the face of the subject can be arranged in an ideal state in the captured image.
[0011]
DETAILED DESCRIPTION OF THE INVENTION
Hereinafter, embodiments of the present invention will be described with reference to the drawings.
[0012]
FIG. 1 schematically shows an external configuration of a human image photographing apparatus according to the present embodiment. This person image photographing device includes an illuminator 2 that illuminates a person to be photographed 1, a CCD type video camera 3 as photographing means for photographing an upper body image including at least a face of the person to be photographed 1, a photographing state and a photographer ( An operator) a display (monitor display) 5 that displays the operation information of the operator 4, and an input image from the camera 3 is processed according to an operation instruction from the photographer 4, or the display 5 or a built-in storage unit A control table (control unit) 6 that performs output, a chair 7 on which the person to be photographed 1 sits when photographing, and a background plate 8 that becomes the background of the person to be photographed 1 when photographing. Here, it is assumed that the color of the background plate 8 is set to blue, for example.
[0013]
FIG. 2 is a block diagram showing an overall configuration of the person image photographing device shown in FIG. That is, an analog signal (input image) from the camera 3 is converted into a digital signal in real time and transferred to a video memory, a main memory or the like, and image data of the video capture unit 11 and other drawing processes are performed. The video accelerator 12 for controlling the display on the display 5, the CPU (Central Processing Unit) for performing various arithmetic processes and various controls on the image data and others, the image data and the calculation results are temporarily stored. Main memory to be stored, CPU unit 13 having an interface with various peripheral devices, external bus control unit 14 for controlling an external bus (for example, SCSI) connecting various peripheral devices, various OSs and applications, control data, Hard disk device that stores image data, etc. A mass storage device 15, a reader / writer unit 17 for reading / writing data from / to a portable storage medium (for example, a magneto-optical disk) 16 for storing output data of the control stand 6, and an operator (photographer) The keyboard 18 and mouse 19 as input devices operated by 4, the image printing device 20 as output means for printing out, for example, a captured face image, and other various devices (image printing device and image on the same network 21). A network adapter (LAN card or the like) 22 for performing mutual communication with a filing device or the like.
[0014]
FIG. 3 is a conceptual diagram showing a head search method used when detecting the top position of the subject 1.
[0015]
The processing operation of the present apparatus will be described below with reference to the flowcharts shown in FIGS. 4 and 5 and FIG. Note that the processing described below is mainly executed under the control of the CPU in the CPU unit 13.
[0016]
First, detection of the top position will be described.
[0017]
In step S <b> 1, an input image 31 (see FIG. 3, hereinafter also referred to as a captured image) obtained from the camera 3 obtained via the video capture unit 11 is captured and temporarily stored in the main memory in the CPU unit 13. Then, according to the input image 31 in the main memory, the color component luminance indicating the high luminance of a total of four points (for example, a, b, c, d in FIG. 3) at the two left and right ends of the upper end of the background image portion. In this case, the blue component luminance (hereinafter referred to as “B luminance”) and the color component luminance indicating low luminance, in this case, the red component luminance (hereinafter referred to as “R luminance”) are obtained, and these four points a , B, c, and d, the average values of B luminance and R luminance are calculated, and the calculated average values are set as reference values bstd1 and rstd1 for the blue background (blue which is the color of the background plate 8). Here, the processing in step S1 corresponds to the first luminance calculation means in the present invention.
[0018]
Next, in steps S2 and S3, the scanning line (search line) is scanned in the horizontal direction sequentially from the upper end in the range (initial detection range) from the upper end line of the captured image 31 to the predetermined line YR. On the other hand, the top position is detected based on the B luminance average value by the binary search method described below. Here, the processing in steps S2 and S3 corresponds to the first top position detecting means in the present invention.
[0019]
1. In step S1, by examining the B luminance at every predetermined pixel from the left end of the captured image 31 within a certain range of the horizontal scanning line (for example, from the X start line to the X end line shown in FIG. 3). The number of B luminance pixels that are not within the obtained B luminance average value bstd1 ± α (α is an error) is counted.
[0020]
2. A scanning line in which the count value in 1 exceeds a preset reference value bnmax1 is a scanning line where the head exists (in the example of FIG. 3, the third search line and the nth search line).
[0021]
3. In the case where the head is present in the scanning line examined in 2 above, the scanning line located at a half distance of the detection range above the scanning line is examined, and when the head does not exist in the scanning line, The scanning line located at half the distance of the detection range below the scanning line is examined. In this way, the head position is narrowed down (detected) by repeating the line investigation with the detection range halved.
[0022]
In this way, when the top position is determined on the basis of the B luminance, the process proceeds to step S4, and as a noise detection process, a scan line where the head is present is predetermined in a downward direction from the scan line where the top position is determined. It is confirmed that there are more than YN lines.
[0023]
In Steps S2 and S3, when the head position cannot be detected based on the B luminance, the process proceeds to Steps S5 and S6, and from the upper end in the range (initial detection range) from the upper end line of the captured image 31 to the predetermined line YR. By sequentially scanning in the horizontal direction, the top position is detected for the scanning line (search line) based on the R luminance average value by the binary search method similar to steps S2 and S3. However, the reference value of R luminance is rstd1 ± β (β is an error), and the maximum count is rnmax1. Here, the processing in steps S5 and S6 corresponds to the second top position detecting means in the present invention.
[0024]
In this way, when the vertex position is determined on the basis of the R luminance, the process proceeds to step S4, and noise detection processing is performed. If the head position is not determined on the basis of the R luminance, the process proceeds to step S22, error processing is performed, and the photographing process is terminated. Here, the processing in step S22 corresponds to the error processing means in the present invention.
[0025]
Note that the error processing in step S22 is, for example, storing the error content in the main memory in the CPU unit 13 so that it can be referred to later, and displaying the error content on the display 5 so that the photographer Guide to 4.
[0026]
Next, detection of the left and right head side positions will be described.
[0027]
First, the left head side position is detected.
[0028]
In step S7, the imaged image 31 stored in the main memory in the CPU unit 13 is added to each of the two left and right edges of the background image portion in the vicinity of the vertex position detected as described above. Color component luminance indicating high luminance of four points, in this case, blue component luminance (hereinafter referred to as B luminance), and color component luminance indicating low luminance, in this case, red component luminance (hereinafter referred to as R luminance) And the average values of the B luminance and R luminance of the four points thus obtained are calculated, and the calculated average values are used as reference values bstd2, rstd2 for the blue background (blue which is the color of the background plate 8). To do. Here, the processing in step S7 corresponds to the second luminance calculation means in the present invention.
[0029]
Next, in steps S8 and S9, in the range (initial detection range) from the left end line of the captured image 31 to the predetermined line XR1, the scanning line (search line) is sequentially scanned from the left end in the vertical direction. On the other hand, the left side position is detected based on the B luminance average value by the binary search method described below. Here, the processes in steps S8 and S9 correspond to the first head side position detecting means in the present invention.
[0030]
1. In a certain range of the vertical scanning line (for example, from the top line to the Y end line shown in FIG. 3), the B luminance is obtained at every predetermined pixel from the upper end of the photographed image 31, and is obtained in step S7. The number of B luminance pixels that are not within the B luminance average value bstd2 ± α (α is an error) is counted.
[0031]
2. A scanning line in which the count value in 1 exceeds a preset reference value bnmax2 is defined as a scanning line in which the head is present.
[0032]
3. In the case where the head is present in the scanning line examined in 2 above, the scanning line located at a half distance of the detection range in the left direction from the scanning line is examined, and the head is not present in the scanning line. The scanning line located at a distance half the detection range in the right direction from the scanning line is examined. In this way, the left-side position is narrowed (detected) by repeating the line investigation with the detection range being halved.
[0033]
When the left head side position is determined on the basis of the B luminance in this way, the process proceeds to step S10, and as a noise detection process, the head is present in the right direction with respect to the scanning line in which the left head side position is determined. It is confirmed that there are at least a predetermined number of scanning lines XN1.
[0034]
In Steps S8 and S9, when the left head side position cannot be detected based on the B luminance, the process proceeds to Steps S11 and S12, and in the range (initial detection range) from the left end line of the captured image 31 to the predetermined line XR1. By sequentially scanning in the vertical direction from the left end, the left head side position is detected for the scanning line (search line) by using the binary search method similar to steps S8 and S9 on the basis of the R luminance average value. . However, the reference value of R luminance is rstd2 ± β (β is an error), and the maximum count is rnmax2. Here, the processing of steps S11 and S12 corresponds to the second head side position detecting means in the present invention.
[0035]
In this way, when the left head position is determined on the basis of the R luminance, the process proceeds to step S10, and noise detection processing is performed. When the left head side position is not determined on the basis of the R luminance, the process proceeds to step S22, error processing is performed, and the photographing process is ended.
[0036]
Next, the right head side position is detected.
[0037]
In steps S13 and S14, in the range from the right end line of the captured image 31 to the predetermined line XR2 (initial detection range), the scanning line (search line) is sequentially scanned from the right end in the vertical direction. The right-side position is detected based on the B luminance average value by the binary search method described below. Here, the processes in steps S13 and S14 correspond to the first head side position detecting means in the present invention.
[0038]
1. In a certain range of the vertical scanning line (for example, from the top line to the Y end line shown in FIG. 3), the B luminance is obtained at every predetermined pixel from the upper end of the photographed image 31, and is obtained in step S7. The number of B luminance pixels that are not within the B luminance average value bstd2 ± α (α is an error) is counted.
[0039]
2. A scanning line in which the count value in 1 exceeds a preset reference value bnmax2 is defined as a scanning line in which the head is present.
[0040]
3. In the case where the head is present in the scanning line examined in 2 above, the scanning line located at half the distance of the detection range in the right direction from the scanning line is examined, and when the head does not exist in the scanning line The scanning line located at a distance half the detection range in the left direction from the scanning line is examined. As described above, the position of the right head side is narrowed (detected) by repeating the line investigation with the detection range being halved.
[0041]
When the right head side position is determined on the basis of the B luminance in this way, the process proceeds to step S15, and as a noise detection process, the head exists in the left direction from the scanning line in which the right head side position is determined. It is confirmed that there are at least a predetermined number of scanning lines XN1.
[0042]
In Steps S13 and S14, when the right position cannot be detected based on the B luminance, the process proceeds to Steps S16 and S17, and in the range from the right end line of the captured image 31 to the predetermined line XR2 (initial detection range) from the right end. By sequentially scanning in the vertical direction, the right head side position is detected for the scanning line (search line) by the binary search method similar to steps S13 and S14 based on the R luminance average value. However, the reference value of R luminance is rstd2 ± β (β is an error), and the maximum count is rnmax2. Here, the processes in steps S16 and S17 correspond to the second head side position detecting means in the present invention.
[0043]
In this way, when the right head position is determined on the basis of the R luminance, the process proceeds to step S15, and noise detection processing is performed. When the right head side position is not determined on the basis of the R luminance, the process proceeds to step S22, error processing is performed, and the photographing process is terminated.
[0044]
For convenience of explanation, it has been described that the position of the left and right heads is detected after the position of the head is detected, but in reality, the detection of the position of the head and the detection of the positions of the left and right heads are performed in parallel.
[0045]
When the detection of the vertex position and the left and right head side positions is completed, the process proceeds to step S18, the center position of the detected left and right head side positions is calculated, and is set as the face center position CC. Here, the processing in step S18 corresponds to the face center position calculating means in the present invention.
[0046]
Next, in step S19, it is confirmed whether or not the vertex position and face center position CC detected previously are within a preset reference range. If the result of this confirmation is that it is not within the reference range, the process proceeds to step S22, error processing is performed, and the photographing process is terminated. Here, the process of step S19 corresponds to the determination means in the present invention.
[0047]
As a result of the confirmation, if it is within the reference range, the process proceeds to step S20, and from the captured image 31 stored in the main memory in the CPU unit 13 based on the head position and the left and right head positions detected previously. The upper body image of the subject 1 is cut out. An example of the cut out upper body image is indicated by reference numeral 32 in FIG. Here, the processing in step S20 corresponds to the upper body cutting means in the present invention.
[0048]
Next, in step S21, the clipped upper body image 32 is converted for output, and together with other personal information of the person to be photographed 1, the image printing apparatus 20 connected to the CPU unit 13 according to the purpose, or In an image printing apparatus (not shown) connected from the network adapter 22 to the network 21, it is printed out on a recording medium (paper, card, etc.) or image filing as an image storage means connected on the network 21. Save and output to a device (not shown), or store the main memory in the CPU 13 or the large-capacity storage device 15 or the portable storage medium 16 as an image storage means. (Save output). This completes all the photographing processes for one person to be photographed.
[0049]
It should be noted that whether the photographed face image is printed out or stored is selected depending on the purpose, and the method of outputting the face image is arbitrary.
[0050]
With the head position detection algorithm as described above, the head position of the subject can be detected with high accuracy even when the subject 1 is not black hair. As an example, the RGB values of black hair, white hair, headless hair, and the head position of the subject are as shown in Table 1 below.
[0051]
[Table 1]

[0052]
As is apparent from Table 1 above, since the B brightness is far from the B brightness of the blue background, the head position can be accurately detected only with the B brightness reference. In addition, because the brightness of white hair and headless hair and B brightness are almost the same as the B brightness of the blue background, it cannot be detected on the basis of the B brightness, but the R brightness is far from the R brightness of the blue background, so It can be seen that the position of the part can be detected with high accuracy.
[0053]
In the above-described embodiment, the case where a blue background is used as the background of the subject and the head position of the subject is detected based on the B luminance and the R luminance of the blue background has been described. For example, another specific color such as red or green may be used as the background of the subject. In this case, the head position of the subject is detected based on the color component luminance indicating high luminance and the color component luminance indicating low luminance for the specific color that is the background of the subject.
[0054]
【The invention's effect】
As described above in detail, according to the present invention, even when the subject is not black hair, the head position of the subject can be accurately detected, and the face position of the subject is ideal in the captured image. A person image photographing method and a person image photographing apparatus that can be arranged so as to be in a state can be provided.
[Brief description of the drawings]
FIG. 1 is a configuration diagram schematically showing an external configuration of a human image photographing apparatus according to an embodiment of the present invention.
FIG. 2 is a block diagram showing an overall configuration of the person image photographing device shown in FIG. 1;
FIG. 3 is a conceptual diagram showing a head search method used when detecting the top position of a subject.
FIG. 4 is a flowchart for explaining a processing operation.
FIG. 5 is a flowchart for explaining a processing operation;
[Explanation of symbols]
1 ... Subject 2 ... Illuminator 3 ... Video camera (photographing means)
4 ... Photographer (operator)
5 …… Display (monitor display)
6 …… Control stand (control unit)
DESCRIPTION OF SYMBOLS 11 ... Video capture part 12 ... Video acceleration part 13 ... CPU part 15 ... Mass storage medium 16 ... Storage medium 17 ... Reader / writer part 20 ... Image printing apparatus (output means)

Claims (4)

A first step of photographing an image including at least a face of a subject with a background of a specific color;
A second step of obtaining a color component luminance indicating a high luminance and a color component luminance indicating a low luminance with respect to a specific color as a background of the subject from the image captured in the first step;
A third step of detecting the head position of the subject based on the color component luminance determined in the second step by a binary search method ;
A fourth step of cutting out an image of the predetermined area of the subject from the image photographed by the first step based on the head position of the subject detected by the third step;
A fifth step of outputting an image of a predetermined region of the person to be photographed cut out by the fourth step as a photographic image,
A person image photographing method characterized by comprising:   The specific color of the background is blue, the color component luminance indicating high luminance with respect to the specific color is blue component luminance, and the color component luminance indicating low luminance with respect to the specific color is red component luminance. The person image photographing method according to claim 1. Photographing means for photographing an image including at least the face of the subject on the background of the specific color;
Luminance calculation means for obtaining a color component brightness indicating a high brightness and a color component brightness indicating a low brightness with respect to a specific color as a background of the subject from the image photographed by the photographing means ;
A head position detecting means for detecting the head position of the subject with reference to the color component brightness obtained by the brightness calculating means by a binary search method ;
Image cutting means for cutting out an image of a predetermined area of the subject from the image taken by the photographing means based on the head position of the subject detected by the head position detecting means;
An output means for outputting an image of a predetermined area of the photographed person cut out by the image cutting means as a taken image;
A person image photographing device characterized by comprising:   The specific color of the background is blue, the color component luminance indicating high luminance with respect to the specific color is blue component luminance, and the color component luminance indicating low luminance with respect to the specific color is red component luminance. The person image photographing device according to claim 3.

Images