KR100220837B1 - Apparatus and method of target tracking in image telephone - Google Patents

Abstract

The present invention relates to a target tracking method and apparatus for changing a direction of a camera by arbitrarily setting a shape of a correlation area using a differentiator and a scanner to track a target and calculating a correlation degree.

The present invention extracts a boundary pixel from an initial input image, extracts a face shape of a talker using a scan method, sets it as a correlation area, and calculates a correlation to reduce a calculation amount and improve a tracking performance.

Therefore, the present invention estimates the correlation by determining the shape of the window through boundary tracking and scanning instead of the conventional square window, and thus it is effective to reduce the calculation amount while improving the tracking performance by tracking the target.

Description

A method and apparatus for tracking a video telephone device (Target Tracking Method and Device for Video phone)

The present invention relates to a target tracking method and apparatus for changing a direction of a camera by arbitrarily setting a shape of a correlation area using a differentiator and a scanner to track a target and calculating a correlation degree.

In general, tracking methods for tracking an object, ie, a target, include a center point tracking method and a correlation tracking method.

The center point tracking method is a method of tracking the center point A of the moving object after separating the moving object from the background as shown in FIG. 1A. At this time, the threshold value is used to separate the moving object, that is, the target, from the background. That is, the background and the object are binarized using the threshold value.

However, this central tracking technique has a disadvantage that it is highly affected by noise.

That is, if the image is relatively simple and the image area is easy and the restriction on the speed of the trackable object is relatively small, the tracking stability is good and the influence of noise is small. On the other hand, if the image is relatively complicated and the image area is not easy and the constraint on the speed of the trackable object is relatively large, the tracking stability is bad and the noise is increased.

1B, the correlation tracking method defines a moving object of a previous frame, that is, a region B having a size appropriate to the position of the target, and calculates a correlation (B) between the defined region B and the search region in the current frame And estimates that the object has moved to the region B 'having the highest degree of correlation.

That is, the correlation tracking method defines a window region of a fixed size including a moving object, that is, a correlation region when a position of a moving object is given in a given n-th image, The position of the region with the highest degree of correlation is regarded as the position of the moving object in the (n + 1) -th image.

Here, in calculating the correlation, the area of the initial window, that is, the area used in calculating the correlation between the current frame and the previous frame, is mainly a square window.

Therefore, the correlation tracking method calculates the correlation directly from the image of the current frame without performing the image segmentation process, so that the tracking performance is maintained even for a relatively complicated image, but the calculation amount is increased.

In other words, since the correlation tracking method generally uses the intensity information of the image without binarizing the image as compared with the center-point tracking method, it is expected that some degree of tracking performance can be expected even when the image area can not be added by adding background scattering. However, the correlation tracking method has a disadvantage in that the computation amount is increased because the correlation degree must be calculated for all cases between the correlation region and the search region in order to estimate the movement of the moving object.

In order to solve the above-mentioned disadvantages, the present invention provides a picture telephone in which boundary pixels are extracted from an initial input picture, a face shape of a talker is extracted using a scanning method, and a correlation area is set as a correlation area, And an object tracking method and apparatus for improving performance.

1A is a view for explaining a conventional center point tracking technique;

1B is a diagram for explaining a conventional correlation tracking technique;

2 is a flowchart of a target tracking method according to the present invention.

FIG. 3A is a view for explaining the initial window setting step of FIG. 2; FIG.

FIG. 3B is a view for explaining the boundary pixel extraction step of FIG. 2; FIG.

FIG. 4 is a view for explaining the scanning operation of FIG. 2; FIG.

5 is a view for explaining the region extracting step of Fig. 2

6 is a view for explaining the correlation degree calculation step of FIG. 2; FIG.

7 is a block diagram of a target tracking apparatus according to the present invention

DESCRIPTION OF THE REFERENCE NUMERALS

300: infrared camera 310: ordinary camera

400: motor section 410: motor driver

420: motor 500: control unit

600: Image storage unit 610: Multiplexer

620: A / D converter 630,640,650,660: memory

700: correlation area shape determination unit 710: boundary pixel extraction unit

720: Scanner 730: Area shape determination unit

800: Correlation calculation unit 900: Motion estimation unit

According to another aspect of the present invention, there is provided a method for tracking a target object of a video phone, the method comprising: selecting an input video signal to be used for tracking in a tracking mode; Determining a shape of a correlation area by extracting a shape of a target through a scan from an input initial image; Extracting a correlation area in which a target is located in a previous image input according to the determined shape of the correlation area and extracting a search area for search from the input current image; A motion estimation and motor driving step of estimating motion according to a correlation between the extracted correlation area and the search area to move the camera; And a call termination step of repeating the region extracting step, the motion estimation, and the motor driving step for an image input until the call is terminated.

According to another aspect of the present invention, there is provided an apparatus for tracking a target object of a video telephone, comprising: video storage means for selecting a video signal to be used for tracking in a tracking mode, A / D converting the video signal and storing the video signal; A correlation area shape determination unit for extracting a shape of a target through scanning from an initial image stored in the image storage unit to determine a shape of a correlation area; A correlation calculation means for extracting a correlation region and a search region from a previous image and a current image output from the image storage means according to the shape of the correlation region determined by the correlation region shape determination means and calculating a correlation degree; A motion estimation means for estimating a motion according to the correlation calculated by the correlation calculation means; And control means for controlling the operation of the image storage means and moving the camera in accordance with the motion estimated by the motion estimation means.

Hereinafter, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

The target tracking method of a video telephone according to the present invention comprises the steps of selecting input video signals 100 and 101, determining a correlation area shape 102, 103, 104, 105 and 106, 107, 108, motion estimation and motor drive steps 109, 110, 111, and call termination steps 110, 111.

The input video signal selection step (100, 101) selects a video signal to be used for tracking in a tracking mode.

In the correlation area shape determination step (102, 103, 104, 105, and 106), the shape of the correlation area is determined by extracting the shape of the target through the scan from the input initial image.

The correlation area shape determination step 102, 103, 104, 105, and 106 may include an initial image storage step 102 for A / D (Analog / Digital) conversion of an input initial image, An initial window setting step (103) for setting an initial window at a predetermined initial window, a boundary pixel extracting step (104) for extracting boundary pixels in the set initial window, a step of scanning in the horizontal and vertical directions along the extracted boundary pixel And a shape determination step (106) of determining the shape of the extracted target mule as the shape of the correlation area.

The initial image is an image taken by the camera when the target is at a predetermined position, and the boundary pixel compares the magnitude of the horizontal and vertical differential values of the brightness values of the pixels in the set initial window with the threshold value for setting the boundary line Lt; / RTI >

In the region extracting step (107, 108), a correlation area in which a target is located is extracted from a previous image input according to the determined shape of the correlation area, a search area for searching is extracted from the input current image, An image storing step 107 for storing A / D (Analog / Digital) converted and stored, a correlation area setting step of extracting a part having the highest correlation extracted from the previous image in the shape of the determined correlation area, (108), and a search region extraction step (108) for extracting a search region from the stored current image with reference to the correlation region.

Here, the search area is four times the size of the rectangle including the correlation area extracted from the previous image.

In the motion estimation and motor driving steps 109, 110 and 111, the motion is estimated according to the correlation between the extracted correlation area and the search area, and the camera is moved.

Here, the correlation function for calculating the correlation is MAE (Mean Absolute Error).

In the call termination step (110, 111), the region extracting step, the motion estimation, and the motor driving step are repeated for the image input until the call is terminated.

A method for tracking a target of a video telephone according to the present invention will be described in detail with reference to the accompanying drawings.

First, the input video signal selection step (100, 101) is performed to select a video signal to be used for tracking in a tracking mode.

Since there may be times when a call is initiated and the user does not want to use the tracking function depending on the situation, the user first decides whether to use the tracking function (100). The distinction between trace mode and non-trace mode can be implemented simply with a switch. That is, if the switch is ON, the tracking function is performed. If the switch is OFF, the tracking function is not performed.

When the tracking mode in which the switch is turned on performs the tracking function, it is determined whether the video signal input from the camera is an infrared signal or a general video signal and selects a video signal to be used for tracking.

That is, the video signal to be used for tracking may be an infrared signal output from an infrared camera or a general video signal output from a general camera. At this time, it is necessary to determine which video signal is to be selected, and to select a desired video signal to track the target.

On the other hand, a general video signal refers to a video signal acquired by a general camera, that is, a camera that acquires brightness information.

After selecting a video signal to be used for target tracking in this way, an initial image storage step 102 is performed in which an initial image is A / D (Analog / Digital) converted and stored.

That is, the video signal coming from the camera is outputted as an 8-bit digital signal through the A / D converter. At this time, a high-bit A / D converter may be used. Since this digital signal is a video signal, it is stored as a digital image I (x, y) having a value within a specified range on the two-dimensional matrix on the memory.

This digital image is an image with a size of 64 x 64 pixels, which is stored in memory and used for target tracking.

In the correlation tracking, it is assumed that the initial position, that is, the position where the target caller can enjoy the dial to make a call, is known by the locking process of the initial target. Thus, in this initial position, the initial image taken by the camera can be obtained when the target is at a preset position for the call.

An initial window setting step 103 is performed to set a window composed of a square of 8 x 8 pixels, which can include a target person's face, as an initial window from the initial image thus taken.

That is, as shown in FIG. 3A, an initial window C consisting of 8 x 8 pixels including a face of the talker is set from an initial image of 64 x 64 pixels.

The boundary pixels are extracted through the differentiator of each pixel in the set initial window, and this process will be described in detail with reference to FIGS. 3B and 3C.

The image of the set initial window is a form in which brightness values are assigned to the two-dimensional matrix I (x, y).

The video signal of the initial window is differentiated as shown in the following equations (1) and (2) in the horizontal and vertical directions.

[Formula 1]

I (x + 1, y-1) + I (x-1, y) + I -I (x + 1, y-1) = h

[Formula 2]

I (x, y + 1) -I (x, y + 1) -I (x + 1, y + 1) = v

As shown in FIG. 3B, the horizontal differential includes the brightness values A4, A5, and A6 of the left pixel and the brightness values A1, A2, and A3 of the right pixel with the reference pixel x, A7 and A1 on the upper side and the brightness values A6, A8 and A3 on the lower side are centered on the reference pixel x and y, It is done by retrieving the car.

As shown in the above equation (1), the horizontal derivative is computed to calculate the horizontal derivative (h), and the vertical derivative is computed as shown in equation (2) h, v) (h ² + v ² ) ^1/2 . That is, the magnitude ((h ² + v ² ) ^1/2 ) of the differential value is calculated by squaring the vertical and horizontal differential values (h, v) and then taking the square root again.

The calculated differential magnitude ((h ² + v ² ) ^1/2 ) is compared with a preset boundary line discrimination threshold value, and when the threshold value is greater than the boundary line discrimination threshold value, the reference pixel (x, y) becomes a boundary pixel.

Such a differentiation process is performed on all the pixels in the initial window, i.e., 8 x 8 pixels.

After the boundary pixels are extracted in this manner, a target shape extraction step 105 is performed to scan the target pixels in the horizontal and vertical directions along the extracted boundary pixels to extract the shape of the target.

In other words, as shown in FIG. 4, when scanning is performed horizontally and vertically, a transition is made from '0' to '1' or '1' to '0' 1 'is extracted and the shape of the target is extracted.

When the shape of the target is extracted, a shape determination step 106 is performed to determine the shape of the correlation area.

That is, as shown in FIG. 4, the shape of the target extracted through the scan is determined as the shape of the correlation region, and the correlation region and the search region are extracted according to the determined correlation region.

First, an input current image is A / D (Analog / Digital) converted and stored (107).

After detecting the shape of the correlation area for the first time in the initial window, since the previous image is the initial image, the extracted correlation area is set as the correlation area as it is. This is because the portion having the highest degree of correlation extracted from the previous image is the extracted correlation region.

After extracting the correlation area, a search area having a size four times the size of the rectangle including the extracted correlation area is extracted from the stored current image (108). That is, as shown in FIG. 5, a search area having a size four times the size of the rectangle including the correlation area in the previous image, i.e., the initial image, is extracted.

The degree of correlation between the extracted retrieved region and the correlated region is calculated (109.

In order to calculate the correlation between the correlation region of the initial image and the search region of the current image, a correlation function to be used must be determined.

Correlation functions for calculating the correlation include NCCF (Normalized Cross Correlation Function), MSE (Mean Square Error), and MAE (Mean Absolute Error).

NCCF (p, q) = [ ΣI n (i, j) · I n + 1 (i + p, j + q)] / [(ΣI n 2 (i, j)) 1/2 · (ΣI n + ₁ ² (i + p, j + q)) ^1/2 ]

MSE (p, q) = E ([I n (i, j) - I n + 1 (i + p, j + q)] 2)

MAE (p, q) = E (┃I n (i, j) - I n + 1 (i + p, j + q) ┃)

Where E (·) represents the mean.

As the correlation function, NCCF is most suitable but MAE is used considering the calculation amount.

The higher the correlation between images, the smaller the value of MAE is. Therefore, when MAE is used as a correlation function, the position with the lowest value becomes the moving object in the next image, that is, the estimated position of the target.

That is, as shown in FIG. 6, the correlation between the search area of the current image and the correlation area of the initial image, which is the previous image, is calculated according to the correlation function, and the part with the highest correlation MAE is the Since it can be seen as a target, the motion is calculated by the difference between the previous position and the current position.

That is, the difference between the position of the correlation area of the initial image, which is the previous image, and the position of the part having the highest correlation in the search area, is calculated to estimate the movement of the face of the target, that is, the talker as shown in FIG. 6 110).

The motor is driven according to the estimated movement, that is, the talker moves to move the camera (111).

At this time, if a new image is inputted from the camera, a new image is inputted, and then an A / D conversion is performed and then an image storing step 107 is performed.

The video signal of one frame is stored in one memory and then shifted to the next memory and used as a previous video. That is, the video signals of the current video and the previous video are stored in the two memories, respectively.

Accordingly, the image stored in the image storage step 107 is the current image to be currently input, and the previous image including the initial image is shifted and stored in another memory for the next motion estimation.

When the video signals of the previous video and the current video are stored in the memory, the correlation area and the search area are extracted (108).

That is, a portion having the highest degree of correlation extracted from the motion estimation step 110, which is the highest degree of correlation extracted from the previous image, is extracted as a shape of the determined correlation region and set as a correlation region, And a rectangle having a size of four times the size of the rectangle including the rectangle (108).

Correlation calculation step 109, motion estimation step 110, and motor drive step 111, which then calculate the correlation, are performed as described above.

On the other hand, if a new image is not input from the camera, a call termination step of terminating the tracking function is performed by searching whether the call is terminated.

That is, if a new image is input, the process goes to the image storage step 107 and the above process is repeated. If the new image is not input and the call is terminated, the tracking function is terminated (112, 113).

If the call is not terminated but a new image is not input, the system waits until a new image is input.

Next, a target tracking apparatus for a video telephone according to the present invention includes an image storage unit 600, a correlation area shape determination unit 700, a correlation degree calculation unit 800, a motion estimation unit 900, a control unit 500, Unit 400, and cameras 300 and 310.

The image storage unit 600 selects a video signal to be used for tracking in a tracking mode and A / D-converts the video signal to be stored in the video recording unit 600, and selects a video signal to be used for tracking in a tracking mode under the control of the controller 500 A multiplexer 610, an A / D converter 620 for A / D-converting the video signal output from the multiplexer 610, and an A / D converter 610 for controlling the A / (630, 650) for respectively storing the current video signals outputted from the correlation calculator (620) and outputting them to the correlation calculator (800); And memories 640 and 660 for respectively storing previous video signals output from the memories 630 and 650 and outputting the previous video signals to the correlation area shape determination unit 700. [

The correlation area shape determination unit 700 determines the shape of the correlation area by extracting the shape of the target through the scan from the initial image stored in the image storage unit 600, A boundary pixel extraction unit 710 for extracting boundary pixels in an initial window at a position of a target in the image, and a shape extracting unit 710 for extracting a shape of the target by scanning in the horizontal and vertical directions according to the boundary pixels extracted by the boundary pixel extraction unit 710 And a region shape determination unit 730 that determines the shape of the target mule extracted from the scan unit 720 as the shape of the correlation region.

The initial image is an image taken by the camera when the target is at a predetermined position, and the boundary pixel compares the magnitude of the horizontal and vertical differential values of the brightness values of the pixels in the set initial window with the threshold value for setting the boundary line Lt; / RTI >

The correlation calculation unit 800 extracts a correlation region and a search region from a previous image and a current image output from the image storage unit 600 according to the shape of the correlation region determined by the correlation region shape determination unit 700 Calculate the correlation.

The search area is four times the size of a rectangle including the correlation area extracted from the previous image, and the correlation function for calculating the correlation is MAE (Mean Absolute Error).

The motion estimation unit 900 estimates motion according to the degree of correlation calculated by the correlation calculation unit 800. [

The control unit 500 controls the operation of the image storage unit 600 and moves the camera according to the motion estimated by the motion estimation unit 900.

The motor unit 400 includes a motor driver 410 for driving the motor under the control of the controller 500 and a motor 420 for moving the camera 300 by the motor driver 410.

The camera outputs two types of video signals to a general camera 310 for outputting a general image and an infrared camera 300 for outputting an infrared image.

The operation of the target tracking device of the image phone according to the present invention constructed as above will be described.

When the tracking mode is turned on, the multiplexer 610 and the A / D converter 620 are turned on according to the control of the controller 500 to A / D convert the video signals coming from the cameras 300 and 310, .

That is, the control unit 500 controls the multiplexer 610 to select a video signal to be used for tracking by determining whether the video signal input from the camera is an infrared signal or a general video signal.

The CPU 500 determines which signal to select from among the infrared image signal and the general image signal. In determining the shape of the correlation area, the infrared image signal input from the infrared camera 300 is used And when the correlation is calculated by accepting the current input image, a general video signal input from a general camera is selected to be used.

The original image signal thus selected and output from the multiplexer 620 is A / D converted by the A / D converter 620, converted into a digital signal, and then stored in the memories 630 and 650). At this time, if the selected video signal is an infrared video signal, it is stored in the memories 630 and 640, and if it is a general video signal, it is stored in the memories 650 and 660.

That is, the current infrared image signal is stored in the memory 630, the previous infrared image signal is stored in the memory 640, the current general image signal is stored in the memory 650, Is stored.

Accordingly, the control unit 500 turns on the corresponding memory according to the type of the video signal selected by the multiplexer 610. [

An infrared image signal stored in the memory 630 or an initial image which is a general image signal stored in the memory 650 is inputted to the boundary pixel extracting unit 710 and is inputted to an initial After the window is set, boundary pixels are extracted.

That is, referring to FIG. 3B, the horizontal and vertical differential values according to the above equations (1) and (2) are compared with the threshold value for edge discrimination, and if it is larger, the boundary pixels are extracted.

This process is performed on the 8x8 pixels constituting the initial window to extract the boundary pixels in the initial window. Then, as shown in FIG. 4, the extracted boundary pixels are input and the face of the speaker is extracted from the scanner 720 do. That is, the scanners 720 scan in the horizontal and vertical directions to extract the face of the caller, and the area shape determination unit 730 determines the face of the extracted caller as the shape of the correlation area.

In order to calculate the correlation, a general video signal is received from the general camera 310. For this purpose, the control unit 500 controls the multiplexer 610. At this time, the infrared image signal can be continuously used without using the general image signal, and a general image signal can be used to extract the shape of the correlation area.

The initial images stored in the memories 630 and 650 are output to the memories 640 and 660 under the control of the control unit 500 and stored as previous images and the video signals received from the cameras 300 and 310 are transmitted to the control unit 500 D converted by the A / D converter 620 and stored in the memories 630 and 650 in accordance with the control.

As described above, the window of the previous image stored in the memories 630 and 650 is moved. That is, in order to calculate the correlation between the search area in the current image and the correlation area in the previous image, the window of the previous image must be moved to match the correlation area.

In the window of the previous image shifted in this manner, as shown in FIG. 5, the correlation calculation unit 800 extracts the correlation area in the shape of the correlation area determined by the area shape determination unit 730, The search region is extracted for the correlation calculation in the current image outputted from the search units 640 and 660.

The degree of correlation is calculated by the MAE in the correlation calculation unit 800 with the extracted correlation region and the search region.

Therefore, as shown in FIG. 6, the motion estimation unit 900 can detect movement of the search region by the difference between the previous position and the current position because the portion having the highest correlation degree, that is, the smallest MAE, .

The controller 500 controls the motor driver 410 to drive the motor 420 using the motion information. Accordingly, the cameras 300 and 310 move to input a new image to the A / D converter 620, and the controller 500 turns on the A / D converter 620 at this time.

The controller 500 turns on the A / D converter 620 and the memories 630 and 650 and then outputs the motion information to the A / D converter 620 until the motion estimator 900 estimates the motion and outputs the motion information. And the memories 630 and 650 are turned off to prevent the video signals from being input from the cameras 300 and 310. [

Through the above operation, the cameras 300 and 310 are moved according to the position of the target so that the cameras 300 and 310 can track the target.

As described above, according to the present invention, the target tracking method and apparatus of a video telephone determines the shape of a window through border tracking and scanning instead of the conventional square window, thereby estimating the correlation degree. It is possible to reduce the calculation amount.

Claims (14)

An input video signal selection step (100, 101) for selecting a video signal to be used for tracking in a tracking mode; A correlation area shape determination step (102, 103, 104, 105, 106) for extracting a shape of a target through a scan from an input initial image to determine a shape of the correlation area; An area extracting step (107, 108) for extracting a correlation area in which a target is located in a previous image input according to the determined shape of the correlation area and extracting a search area for searching from the input current image; A motion estimation and motor driving step (109, 110, 111) for moving the camera by estimating motion according to the degree of correlation between the extracted correlation area and the search area; And And a call terminating step (110, 111) for repeating the area extracting step, the motion estimation and the motor driving step for the image input until the call is terminated. The method according to claim 1, wherein the initial image is an image taken by a camera when the target is at a predetermined position. 2. The method of claim 1, wherein the determining of the correlation area shape comprises: an initial image storage step (102) for A / D (Analog / Digital) conversion of an input initial image; An initial window setting step (103) of setting an initial window at a position of a target in the stored initial image; A boundary pixel extraction step (104) of extracting boundary pixels in the set initial window; A target shape extracting step (105) of extracting the shape of the target by scanning in the horizontal and vertical directions along the extracted boundary pixels; And determining a shape of the extracted target mule as a shape of a correlated region. 4. The video telephone according to claim 3, wherein the boundary pixels are determined by comparing magnitudes of differential values in the horizontal and vertical directions of the brightness values of the respective pixels in the set initial window with thresholds for setting boundary lines to be set. Way. 2. The method according to claim 1, wherein the search area is four times the size of a rectangle including a correlation area extracted from a previous image. 2. The method of claim 1, wherein the correlation function for calculating the correlation is a Mean Absolute Error (MAE). 6. The method according to claim 5, wherein the region extracting step (107, 108) comprises an image storing step (107) for A / D (Analog / Digital) conversion of the input current image and storing the current image; A correlation area setting step (108) of extracting a part having the highest correlation extracted from the previous image as a shape of the determined correlation area and setting a correlation area; And a search region extracting step (108) of extracting a search region based on the correlation region in the stored current image. An image storage means 600 for selecting a video signal to be used for tracking in the tracking mode and performing A / D conversion and storing the selected video signal; A correlation area shape determination unit (700) for extracting a shape of a target through scanning from an initial image stored in the image storage unit (600) to determine a shape of a correlation area; A correlation calculation unit (700) for calculating a correlation by extracting a correlation area and a search area from a previous image and a current image output from the image storage unit (600) according to the shape of the correlation area determined by the correlation area shape determination unit (800); A motion estimation unit 900 for estimating a motion according to the degree of correlation calculated by the correlation calculation unit 800; And And control means (500) for controlling the operation of the image storage means (600) and controlling the movement of the camera in accordance with the motion estimated by the motion estimation means (900) Device. 9. The apparatus of claim 8, wherein the image storage means (600) A multiplexer (610) for selecting a video signal to be used for tracking in a tracking mode under the control of the control means (500); An A / D converter 620 for A / D-converting the video signal output from the multiplexer 610; D converter 620 according to the type of the video signal under the control of the control means 500 and outputs the current video signals to the correlation calculation means 800. [ Two memories 630, 650; And third and fourth memories 640 and 660 for respectively storing previous video signals output from the first and second memories 630 and 650 and outputting the previous video signals to the correlation area shape determining means 700 Wherein the target tracking unit comprises: 10. The apparatus according to claim 9, wherein the correlation area shape determination means (700) comprises: a boundary pixel extraction unit (710) for extracting boundary pixels in an initial window at a target position in an initial image stored in the image storage means (600); A scan unit 720 for scanning the target pixels in the horizontal and vertical directions according to the boundary pixels extracted by the boundary pixel extracting unit 710; And a region shape determination unit (730) that determines the shape of the target mule extracted from the scaner (720) as a shape of a correlation region. 11. The apparatus according to claim 10, wherein the initial image is a video taken by the camera when the target is at a predetermined position. 11. The video telephone according to claim 10, wherein the boundary pixels are determined by comparing magnitudes of differential values in horizontal and vertical directions of brightness values of the respective pixels in the set initial window with thresholds for setting boundary lines to be set. Device. 9. The apparatus according to claim 8, wherein the search area has a size four times the size of a rectangle including the correlation area extracted from the previous image. 9. The apparatus according to claim 8, wherein the correlation function for calculating the correlation is a Mean Absolute Error (MAE).