KR20130064556A - Apparatus and method for face detection using multi detection - Google Patents

Abstract

PURPOSE: An apparatus for detecting a face using a multiple detection mode and a method thereof are provided to detect a face by extracting a partial area in which a face exists from an input image, and detecting the face around the extracted area, without detecting the whole input image. CONSTITUTION: A face candidate area detection unit(20) extracts a skin color pixel from an image reduced in an image preprocessing unit(10), and extracts at least one skin color area by connecting adjacent skin color pixels, and detects at least one face candidate area on the basis of the extracted skin color area. A face discrimination unit(30) discriminate a final face area using face characteristics information from at least one detected face candidate area. [Reference numerals] (10) Image preprocessing unit; (20) Candidate face region detection unit; (30) Face discrimination unit; (AA) Image

Description

Apparatus and method for face detection using multi detection}

The present invention relates to a face detection method for detecting a face included in an image, and more particularly, by detecting a face candidate region using skin color information and finally detecting the face using facial feature information. The present invention relates to a face detection apparatus and method using a multiple detection scheme capable of face detection.

In general, face recognition technology used in security or user recognition technology is composed of three stages: face detection, feature extraction, and identity verification. In order to improve the performance of the face recognition result, first of all, accurate face detection must be preceded, and for this purpose, various face detection techniques have been developed.

The most widely used method of face detection is to use face texture information. The image to be detected is separated into a set of rectangular regions of arbitrary size where a face can be detected, and the texture information is extracted from each region. The facial region is determined by comparing the extracted texture information with a previously learned facial feature model.

However, in the case of detecting a face using the above-described face texture information, since a face must be detected in a situation in which no prior information on the face size is given, the computational complexity of the algorithm becomes large and the face is often repeatedly executed. There is a disadvantage that the time required to detect is long.

In addition, since the camera of the video security system has been digitized and the image quality is becoming higher in size than SD / HD level, it is suitable for real time processing such as face detection in such high definition video or video and can detect the face at high speed. The need for face detection technology is emerging.

The present invention has been proposed to solve the above-mentioned problems. In particular, a partial region where a face may be present is first detected without performing face detection on the entire image input during face detection, and then the center of the corresponding area is detected. It is an object of the present invention to provide a face detection apparatus and method using a multiple detection method that can detect a face more quickly and reduce the computational complexity of an algorithm by performing face detection.

Another object of the present invention is to provide a face detection apparatus and method using a multi-detection method capable of high-speed face detection for a high-quality image or video.

Face detection apparatus using a multiple detection method of the present invention for achieving the above object is an image preprocessing unit for reducing the size of the image to a predetermined ratio, the skin color pixels after extracting the skin color pixels from the image received through the image preprocessor Extracts at least one skin color region, detects at least one face candidate region based on the extracted skin color region, and detects a final face region using facial feature information from at least one face candidate region. And a face discriminating unit for discriminating.

At this time, the image preprocessor separates the image into a brightness channel and a color channel, reduces the size of the separated color channel at a predetermined ratio, and the face candidate region detector extracts the skin color pixels from the reduced color channel, and then connects adjacent skin color pixels. At least one skin color area is extracted.

Here, the predetermined ratio may be an inverse value of a horizontal or vertical value of the minimum face size at which the face may be detected, and the minimum face size may be set based on an angle or a distance between the camera and the subject photographing the image.

On the other hand, the face candidate region detection unit extracts at least one skin region by connecting adjacent skin color pixels, matches the position coordinates of the skin region with the position coordinates of the corresponding region in the original image, and then resizes the corresponding region in the original image. Will be restored.

The face candidate region detector may extract the face candidate region by masking the skin color region restored to the size of the corresponding region in the original image with the brightness channel of the image.

In addition, the face discriminator extracts a face region from the extracted face candidate region by using face texture information, and then compares the face region with a previously learned face feature model to determine a final face region.

According to an aspect of the present invention, there is provided a face detection method using a multi-detection method to achieve an object as described above. And extracting the final face region by using the facial feature information in the at least one face candidate region.

In this case, the step of reducing the size of the image by a predetermined ratio includes dividing the image into a brightness channel and a color channel, and reducing the size of the separated color channel by a certain ratio, and extracting the face candidate region. Extracting skin color pixels from the reduced color channel, extracting at least one skin color region by connecting the extracted skin color pixels to adjacent skin color pixels, and extracting at least one face candidate region from the at least one skin color region. It may include.

In this case, the separating of the image into the brightness channel and the color channel may be divided into the brightness channel and the color channel using at least one color space conversion function of YCbCr, YUV, CIE LAB, CIE LUV, and HSV.

At this time, the step of reducing the ratio may determine a ratio to be reduced based on the inverse value of the horizontal or vertical value of the minimum face size that can detect the face.

Here, the minimum face size that can detect the face may be determined according to the angle or distance between the camera and the subject which captured the image.

In the extracting of the skin color pixels, a color channel set is selected by selecting a color channel required for skin color extraction from the color channels of the image, and then the skin color pixels are extracted from the color channel set.

In addition, the step of extracting the skin color pixel is determined as the skin color pixel if it is larger than the predetermined threshold value.

In the extracting of the skin color region, at least one skin color region having a quadrangular shape may be extracted by connecting the extracted skin color pixels to adjacent skin color pixels.

The extracting of the face candidate region may include matching the position coordinates of the extracted at least one skin color region with the position coordinates of the corresponding region in the original image, and restoring the size of the corresponding region in the original image. The method may include adjusting a skin color region in consideration of an error range of a predetermined region in the region, and extracting a face candidate region by masking the adjusted at least one skin color region with a brightness channel of the image.

The determining of the final face area may include extracting a face area using face texture information from the extracted face candidate area and determining a final face area by comparing with a previously learned face feature model. .

Here, the face texture information may be face brightness information.

According to the present invention, face detection can be performed more quickly by first extracting a partial area where a face may be present, and then performing face detection centering on the corresponding area without performing detection on the entire input image during face detection. Has the advantage.

According to the present invention, by detecting the skin color region where a face may be present without detecting the entire image input during face detection, the computational complexity is not large and the error at the time of detection can be prevented. There is an advantage.

According to the present invention, it is possible to detect a face more quickly when detecting a face, so that face detection in a high quality image or a real time video can be processed at high speed.

1 is a block diagram showing a main configuration of a face detection apparatus using a multiple detection method according to an embodiment of the present invention.
2 is a schematic flowchart of a face detection method using a multiple detection method according to an exemplary embodiment of the present invention.
3 is a detailed flowchart of a face detection method using a multiple detection method according to an exemplary embodiment of the present invention.
4 is a detailed flowchart of a process of extracting a face candidate region according to an embodiment of the present invention.
5 is an exemplary view showing that the image according to the embodiment of the present invention is reduced to a minimum face size capable of face detection.
6 is an exemplary diagram illustrating a skin color pixel, a skin color region, and a face candidate region according to an embodiment of the present invention.
7 is an exemplary view of a face candidate image extracted according to an embodiment of the present invention.
8 is an exemplary view illustrating a face detection process step by step using a multiple detection method according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the detailed description of known functions and configurations incorporated herein will be omitted when it may unnecessarily obscure the subject matter of the present invention. This is to more clearly communicate without obscure the core of the present invention by omitting unnecessary description. The same reference numerals are used for portions having similar functions and functions throughout the drawings.

Hereinafter, a face detection apparatus using a multiple detection scheme according to an exemplary embodiment of the present invention will be described.

1 is a block diagram showing a main configuration of a face detection apparatus using a multiple detection method according to an embodiment of the present invention.

Referring to FIG. 1, the face detection apparatus 100 using the multiple detection method of the present invention includes an image preprocessor 10, a face candidate region detector 20, and a face discriminator 30.

When the image is input, the image preprocessing unit 10 performs a preprocessing process for detecting a face candidate region from the input image. In particular, the image preprocessing unit 10 includes a brightness channel representing brightness information and a color channel representing color information. The color channels are separated and the size of the separated color channels is reduced at a predetermined ratio.

Here, the predetermined ratio may be obtained as an inverse value of the horizontal or vertical value of the minimum face size in which the face is detected in the input image, and the minimum face size in which the face is detected may be determined by the angle or distance between the camera and the subject which captured the image. Can be set based on

The face candidate region detector 20 extracts a skin color pixel from an image received through the image preprocessor 10, connects adjacent skin color pixels, and extracts at least one skin color region, based on the extracted skin color region. One face candidate region is detected.

In this case, after extracting at least one skin color region by connecting adjacent skin color pixels, matching the position coordinates of the extracted skin color region with the position coordinates of the corresponding region in the original image and restoring the size of the corresponding region in the original image. do. Thereafter, the restored skin color region is masked with the brightness channel separated by the image preprocessing unit 10 described above, and then the face candidate region is extracted.

The face discrimination unit 30 determines the final face region by using face feature information in at least one face candidate region extracted by the face candidate region detection unit 20. At this time, the face discriminator 30 extracts the face region from the face candidate region by using the face texture information and compares the face region with a previously learned face feature model to finally determine the face region.

More detailed operations of the image preprocessor 10, the face candidate region detector 20, and the face discriminator 30 will be described later with reference to FIGS. 2 and 4.

Hereinafter, a face detection method using a multiple detection method in the face detection apparatus 100 will be described.

2 is a schematic flowchart of a face detection method using a multiple detection method according to an exemplary embodiment of the present invention.

1 and 2, in order to detect a face according to an exemplary embodiment of the present invention, an image to be detected must first be input to the image preprocessor 10 (S101). In this case, an image captured by a separate external camera may be input to the face detection apparatus 10 through a communication interface, or may be input by receiving an image from a built-in camera.

Thereafter, when the image is input to the image preprocessor 10, the size of the image is reduced to a predetermined ratio (S103), and then the reduced image is transmitted to the face candidate region detection unit 20. The face candidate region detector 20 extracts the face candidate region from the image received from the image preprocessor 10 (S105), and the face discriminator 30 finally uses the facial feature information in the extracted face candidate region. The face area is determined (S107).

As described above, instead of detecting the face in the entire region when detecting the face using the facial feature information in the image, the face candidate region determined to exist will be first detected based on the reduced image, and the candidate region will be detected. Only by using the facial feature information to determine the final face area, there is an advantage that a high-speed face detection is possible.

Hereinafter, a more specific face detection method will be described with reference to FIG. 3.

3 is a detailed flowchart of a face detection method using a multiple detection method according to an exemplary embodiment of the present invention.

Referring to FIG. 3, when an image is input (S201), the image is first divided into a brightness channel representing brightness information and a color channel representing color information (S203).

)를 이용하여 밝기 채널 및 색상 채널로 분리하게 된다. In this case, the input image may be an RGB image, and a color space conversion function (

) Into the brightness and color channels.

는 밝기 채널을 의미하며,

은

개의 색상 채널 가운데 n번째 색상 채널을 의미한다.here,

Means brightness channel,

silver

The nth color channel of the color channels.

)는 YCbCr, YUV, CIE LAB, CIE LUV, HSV 중 적어도 하나의 색 공간 변환 함수를 이용할 수 있으며, 그 외 색 공간을 표현하는 그 어떤 함수도 적용 가능하다.In this case, the color space conversion function (

) May use at least one of the color space conversion functions of YCbCr, YUV, CIE LAB, CIE LUV, and HSV, and any other function representing the color space is applicable.

Incidentally, here, YCbCr and YUV are composed of luminance information indicating the brightness of light and color difference information (Cb / Cr, U / V), YCbCr is a digital color representation method, and YUV is analog Means the color representation.

In addition, CIE LAB and CIE LUV are to complement the CIE XYZ color space, where the CIE XYZ color space is a color space formed by measuring how a person perceives short, medium, and long wavelength colors. In this case, the accuracy of calculating the color difference between the two colors is disadvantageous.In order to compensate for this, LIE (luminosity) information and color information (A / B, U / V) including CIE LAB, CIE LUV are included. Color space conversion function.

HSV is a color space conversion function expressed as a value of Hue, Saturation, and Brightness, and various color space conversion functions such as YIQ, HLS, and PAL may be applied.

)은 하기의 수학식 2와 같이 영상 크기 조정 함수(

)를 이용하여 일정 비율로 축소된 색상 채널(

)로 조정된다(S205).If the image is separated into a brightness channel and a color channel in step S203, the separated color channel (

) Is an image scaling function (

) To scale color channels (

(S205).

는 영상 크기 조정 함수(

)의 스케일 팩터(scale factor), 다시 말해, 색상 채널을 어떠한 비율에 따라 축소할 지를 나타내는 비율을 의미한다. 보다 구체적으로, S203 단계에서 분리된 색상 채널에서, 얼굴 검출의 속도를 보다 향상시키기 위해, 색상 채널의 크기를 조절하며, 이때, 색상 채널의 크기를 얼굴 검출이 가능한 최소 얼굴 크기의 가로 또는 세로 값의 역 값을 기초로, 색상 채널의 크기를 축소할 수 있다.here,

Is an image scaling function (

) Scale factor, that is, a ratio indicating a ratio of a color channel to be reduced. More specifically, in the color channel separated in step S203, in order to further improve the speed of face detection, the size of the color channel is adjusted, wherein the size of the color channel is the horizontal or vertical value of the minimum face size that can detect the face Based on the inverse value of, the size of the color channel can be reduced.

가 스케일 팩터(

)가 되며, 상술한 과정을 통해 도출된 스케일 팩터(

)는 입력되는 영상의 크기를 조정하는 데 사용되게 된다.More specifically, after extracting at least one or more sample frames including a face image from the input image, the face detection is performed in the at least one extracted sample frame, and the face can be detected in the sample frame based on the detected face size The minimum face size will be determined. For example, if the minimum face size detected in the sample frame is 20x20 pixels, the inverse value of the horizontal or vertical value of the minimum face size may be 1x1.

Is the scale factor (

), And the scale factor derived through the above-described process (

) Will be used to adjust the size of the input image.

)가 결정되면, 상술한 수학식 2에 따라 색상 채널의 크기를 축소하게 되며, 예컨대, 도 5에 도시된 바와 같이 70x70 크기의 영상을 3x3 크기의 영상으로 축소되게 된다.Scale factor

), The size of the color channel is reduced according to Equation 2 described above. For example, as shown in FIG. 5, the 70x70 size image is reduced to the 3x3 size image.

Here, the minimum face size that can detect the face varies depending on the angle and distance between the camera taking the image and the subject (person). More specifically, the size of the face photographed from the camera varies according to a camera setting such as a photographing angle between the camera and the subject (person) or zoom in / zoom out. In addition, even with the same shooting conditions and cameras of the same specification, the size of the face to be photographed will vary depending on the distance between the subject (person) and the camera. Since the size of the face to be photographed is different, the minimum face size that can be detected is different. do.

Thereafter, the skin color pixels are extracted from the color channel reduced in a predetermined ratio (S207).

)을 기초로 피부색 픽셀을 추출하게 되는 데, 이때, 피부색 픽셀 검출 속도를 향상시키기 위해, 모든 색상 채널을 대상으로 피부색 픽셀을 추출하는 것이 아니라, 피부색 추출에 필요한 색상 채널만을 추출하여 피부색 추출을 위한 색상 채널 집합을 구성한 후 추출할 수도 있다.In more detail, the reduced color channel (

Skin color pixels are extracted based on), and in order to improve the speed of skin color pixel detection, instead of extracting skin color pixels for all color channels, only the color channels necessary for skin color extraction are extracted for skin color extraction. You can also extract a set of color channels.

)이 구성되면, 하기의 수학식 3과 같이 피부색 추출 함수(

)을 이용하여 피부색 픽셀 여부를 판별한다.After that, a set of color channels (

) Is configured, the skin color extraction function (

) To determine the color of the skin color pixels.

는 임의의 픽셀 세로 및 가로 위치의 좌표이며,

는 피부색 픽셀을 나타낸다. 또한,

과

은 각각 축소된 영상의 세로와 가로 픽셀 위치의 집합이고,

과

은 각각 축소된 색상 채널의 높이와 넓이를 나타낸다.

는 기설정된 임계값 정보로, 특정 위치의 픽셀이

보다 크다면 해당 픽셀은 피부색 픽셀(

)로 판단하게 된다.here,

Is the coordinate of any pixel vertical and horizontal position,

Represents skin color pixels. Also,

and

Are the set of vertical and horizontal pixel positions of the reduced image, respectively.

and

Denotes the height and width of the reduced color channel, respectively.

Is preset threshold information.

Is greater than the skin color pixel (

Will be judged.

When skin color pixels are extracted in step S207, the skin color region is extracted by connecting adjacent skin color pixels (S209). In this case, before extracting the skin color region, in order to improve detection performance, the image from which the skin color pixels are extracted may be converted into a binary image.

Subsequently, at least one skin color region connecting adjacent skin color pixels is extracted using a function of extracting a skin color region as shown in Equation 4 below.

는 추출된 피부색 영역이며,

는 피부색 영역을 추출하는 함수이고,

은 이진 처리된 피부색 픽셀의 집합을 의미한다.here,

Is the extracted skin color region,

Is a function that extracts the skin color region,

Means a set of binary skin color pixels.

의 값이 파라미터를 의미한다. 여기서

는 임의의 사각형 피부색 영역의 왼쪽 상단 지점의 가로축 좌표이며,

는 임의의 사각형 피부색 영역의 왼쪽 상단 지점의 세로축 좌표이고,

는 임의의 사각형 피부색 영역의 넓이를 의미하며,

는 임의의 사각형 피부색 영역의 높이를 의미한다.In this case, the skin color region connecting the adjacent skin color pixels is preferably configured in a square shape, but is not limited thereto. In order to form a square shape, four parameters are required for each skin color area.

Means a parameter. here

Is the horizontal coordinate of the upper left point of any rectangular skin color area.

Is the vertical coordinate of the upper left point of any rectangular skin color area,

Means the width of any rectangular skin color area,

Means the height of any rectangular skin color area.

Thereafter, the face candidate region is extracted based on the extracted at least one skin color region (S211), and the face region is extracted using face texture information which is facial feature information (S213).

At this time, the facial region is extracted from the extracted face candidate region using facial texture information, for example, face brightness information, and then the final facial region is determined by comparing with the previously learned facial feature model (S215).

In the exemplary embodiment of the present invention, a face region is extracted by using face texture information such as brightness information. For example, the present invention is not limited thereto, and any method may be used as long as it is a face detection method such as an edge detection method. .

A process of extracting a more specific face candidate region will be described with reference to FIG. 4.

As such, instead of extracting the face region from the entire input image, the face detection process is performed only in the region divided into the skin color region, thereby reducing unnecessary face detection process, thereby making it possible to detect the face at high speed. There is an advantage.

In addition, since only the skin color region is included as a face candidate region, an error generated during face detection can be minimized, thereby improving face detection performance.

Hereinafter, a process of extracting a face candidate region will be described in more detail.

4 is a detailed flowchart of a process of extracting a face candidate region according to an embodiment of the present invention.

Referring to FIG. 4, when at least one skin color region is extracted in step S209 of FIG. 3, the extracted skin color region may be re-introduced in the original image to prevent an error in face detection and to further improve detection performance. The size of the corresponding area is restored (S301). At this time, a calculation for restoring the original size is performed as shown in Equation 5 below.

는 복원된 사각형 피부색 영역의 왼쪽 상단 지점의 가로축 좌표이고,

는 복원된 사각형 피부색 영역의 왼쪽 상단 지점의 세로축 좌표이고,

는 복원된 사각형 피부색 영역의 넓이를 나타내며,

는 복원된 사각형 피부색 영역의 높이를 나타낸다.here,

Is the horizontal coordinate of the upper left point of the restored rectangular skin color area.

Is the vertical coordinate of the upper left point of the restored rectangular skin color area.

Represents the area of the restored rectangular skin color area,

Represents the height of the restored rectangular skin color region.

Thereafter, the set of restored skin color regions is generated as in Equation 6 below.

는 원본 영상을 기준으로 복원된 피부색 영역의 집합을 의미한다. here,

Denotes a set of skin color regions reconstructed based on the original image.

Thereafter, in order to further prevent an error in face detection, the skin color region is adjusted in consideration of an error range as shown in Equation 7 below (S303).

는 피부색 영역의 위치 정보인

에서 얼굴 후보 영역을 포함하기 위해 오차 범위를 고려하여 조정된 얼굴 후보 영역의 위치 값을 나타내며,

와

는 각각 피부색 영역의 가로 축과 세로 축의 오차 범위를 구하기 위한 값으로, 예컨대 0보다 크고 1보다 작은 값을 가질 수 있다.here,

Is the location information of the skin color area

Represents a position value of the face candidate region adjusted to include the face candidate region in consideration of an error range.

Wow

Are values for obtaining an error range between the horizontal axis and the vertical axis of the skin color region, respectively, and may have values greater than 0 and less than 1, respectively.

Examples of skin color pixels extracted through the above process, skin color areas connecting adjacent skin color pixels, and face candidate areas adjusted in consideration of an error range of a predetermined area are illustrated in FIGS. 21, 22, and 23 of FIG. 6.

Referring back to FIG. 4, when the face candidate region is extracted through the above-described process (S303), the extracted face candidate region is masked with the brightness channel of the input image (S305) to extract the final face candidate region. (S307).

Since the final face candidate region extraction process has been described with reference to FIG. 3, a detailed description thereof will be omitted.

The face candidate region extracted through the above process is illustrated in FIG. 7.

8 is an exemplary view illustrating a face detection process step by step using a multiple detection method according to an embodiment of the present invention.

Referring to FIG. 8, in the face detection method using the multi-detection method according to the embodiment of the present invention, when an image is input as described above (step a), the image is divided into a brightness channel and a color channel, Reduce the size by a percentage (step b).

Then, after extracting the skin color pixels, and before the adjacent skin color pixels are connected to extract the skin color region, it is preferable to convert to a binary image to improve detection performance as shown in step c.

After extracting the skin color region (d) by connecting adjacent skin color pixels in a binary image, and restoring it to the size of the original image, adjusting the face candidate region in consideration of an error range of a certain region, and brightness Masking is performed with the channel (step e). Then, after extracting the face region using the face texture information (step f), the final face region is determined (g) by comparing with the previously learned face feature model.

As described above, face detection can be performed more quickly by detecting only a partial region where a face may be present without detecting the entire input image.

In addition, by performing detection only on the skin color region where a face may be present, there is an advantage that the computational complexity is not large and errors in detection can be prevented in advance.

While the present invention has been particularly shown and described with reference to exemplary embodiments thereof, it is to be understood that the invention is not limited to the disclosed exemplary embodiments, but, on the contrary, It will be appreciated by those skilled in the art that numerous changes and modifications can be made to the invention. And all such modifications and changes as fall within the scope of the present invention are therefore to be regarded as being within the scope of the present invention.

The present invention relates to a face detection method for detecting a face included in an image, and more particularly, by detecting a face candidate region using skin color information and finally detecting the face using facial feature information. The present invention relates to a face detection apparatus and method using a multiple detection scheme capable of face detection.

According to the present invention, it is possible to detect a face at a high speed by performing detection based on a partial area where a face may exist without detecting the entire input image when detecting a face. It has the effect of contributing to the development of the industrial sector.

10: image preprocessor 20: face candidate region detector 21: skin color pixel
22: skin color region 23: face candidate region 30: face discriminating unit
100: face detection device

Claims (18)

An image preprocessor that reduces the size of the image by a predetermined ratio;
After extracting a skin color pixel from the image received through the image preprocessor, a face candidate region for extracting at least one skin color region by connecting adjacent skin color pixels and detecting at least one face candidate region based on the extracted skin color region Detection unit; And
A face discriminating unit which determines a final face region using face feature information in the at least one face candidate region;
Face detection apparatus using a multiple detection method characterized in that it comprises a. The method of claim 1,
The image preprocessing unit
Separates the image into brightness and color channels, reduces the size of the separated color channels by a percentage,
The face candidate region detection unit
And extracting skin color pixels from the reduced color channel, and extracting at least one skin color region by connecting adjacent skin color pixels. The method of claim 2,
The constant ratio is
Face detection apparatus using a multiple detection method, characterized in that the inverse value of the horizontal or vertical value of the minimum face size capable of face detection. The method of claim 3, wherein
The minimum face size is
Face detection apparatus using a multiple detection method, characterized in that set based on the angle or the distance between the camera and the subject photographed the image. The method of claim 2,
The face candidate region detection unit
Extracting at least one skin color region by connecting the adjacent skin color pixels, matching the position coordinates of the skin color region with the position coordinates of the corresponding region in the original image, and restoring the size of the corresponding region in the original image. A face detection device using a multiple detection method. 6. The method of claim 5,
The face candidate region detection unit
And a face candidate region is extracted by masking the skin color region restored to the size of the corresponding region in the original image with the brightness channel of the image. The method of claim 1,
The face discrimination unit
And extracting a face region from the extracted face candidate region using face texture information, and then comparing the face feature model with a previously learned face feature model to determine a final face region. Reducing the size of the image at a predetermined ratio;
Extracting at least one face candidate region by connecting adjacent skin color pixels in the reduced image; And
Determining a final face region by using face feature information in the at least one face candidate region;
Face detection method using a multiple detection method characterized in that it comprises a. The method of claim 8,
Reducing the size of the image by a predetermined ratio
Separating the image into a brightness channel and a color channel; And
Reducing the size of the separated color channel by a predetermined ratio;
Including;
Extracting the face candidate region
Extracting skin color pixels from the reduced color channel;
Extracting at least one skin color region by connecting the extracted skin color pixels to adjacent skin color pixels; And
Extracting at least one face candidate region from the at least one skin color region;
Face detection method using a multiple detection method characterized in that it comprises a. The method of claim 9,
Separating the image into a brightness channel and a color channel
A face detection method using a multiple detection method, characterized in that the separation into a brightness channel and a color channel using at least one of the color space conversion function of YCbCr, YUV, CIE LAB, CIE LUV, HSV. The method of claim 8,
The step of reducing the ratio is
A face detection method using a multiple detection method, characterized in that for determining the rate of reduction based on the inverse value of the horizontal or vertical value of the minimum face size capable of face detection. 12. The method of claim 11,
The minimum face size at which the face can be detected is
Face detection method using a multiple detection method, characterized in that determined according to the angle or the distance between the camera and the subject photographed the image. The method of claim 9,
Extracting the skin color pixel
Selecting a color channel required for skin color extraction from the color channels of the image to configure a color channel set, and extracting skin color pixels from the color channel set. The method of claim 9,
Extracting the skin color pixel
The face detection method using the multiple detection method, characterized in that it is determined as a skin color pixel when the predetermined threshold value is larger. The method of claim 9,
Extracting the skin color region
A method of detecting a face using a multiple detection method, comprising: extracting at least one skin color region having a quadrangular shape by connecting the extracted skin color pixels to adjacent skin color pixels. The method of claim 9,
Extracting the face candidate region
Matching the position coordinates of the extracted at least one skin color region with the position coordinates of the corresponding region in the original image and restoring the size of the corresponding region in the original image;
Adjusting the skin color area in consideration of an error range of a predetermined area in the restored skin color area; And
Extracting a face candidate region by masking the adjusted at least one skin color region with a brightness channel of an image;
Face detection method using a multiple detection method characterized in that it comprises a. The method of claim 8,
The determining of the final face area may
Extracting a face region from the extracted face candidate region by using face texture information; And
Determining a final face region by comparing with a learned face feature model;
Face detection method using a multiple detection method characterized in that it comprises a. 18. The method of claim 17,
The face texture information
Face detection method using a multiple detection method characterized in that the face brightness information.

Images