KR20180050967A - Device for extracting image - Google Patents

Abstract

The present invention relates to a device for extracting a subject from an image, the device including: a camera unit for photographing a region of a specific place to output an image signal; an outline detecting unit for calculating an outline image from the image photographed by the camera unit; a field memory for receiving the outline image calculated by the outline detecting unit and storing the outline image as a background outline image during one field; and a control unit for calculating a background outline image from an inputted background image and storing the calculated background outline image in the field memory when the background image photographed first by the camera unit is inputted, and then calculating a current outline image from an inputted current image when the current image photographed later by the camera unit is inputted, and then calculating and outputting an output outline image using the calculated current outline image and the stored background outline image.

Description

[0001] DEVICE FOR EXTRACTING IMAGES [0002]

In particular, the present invention relates to an apparatus and method for extracting a subject from an image, and more particularly, to a method and apparatus for extracting a background image from an image and extracting a newly displayed subject image using a pre- And more particularly,

In general, erasing the background from the image requires a lot of time to erase the background from the image because the user manually erases the image by using a utility program such as Photoshop. Therefore, Which causes a great deal of cost in processing.

SUMMARY OF THE INVENTION It is, therefore, an object of the present invention to provide an apparatus and method for identifying a specific region using an image having a low sensitivity to light intensity change or illumination change in an image, and deleting a background by distinguishing a specific region from another region .

An apparatus for erasing a background in an image includes a camera unit for photographing an area of a specific place and outputting the image as a video signal, an outline detection unit for calculating an outline image from the image captured by the camera, and an outline detection unit A field memory for storing a background contour line image in a field for inputting and receiving a background image taken from the camera, a background contour line image is calculated in the input background image, and the calculated background contour line image is stored in a field memory A control unit for calculating a current contour line image from the input current image when a current image photographed later from the camera is inputted and outputting an output contour line image using the calculated current contour line image and the stored background contour line image, .

An apparatus for erasing a background in an image includes a background image generation unit for performing a calculation using a Sobel outline detection mask based on each pixel of the input background image when a background image photographed first from a camera is input to calculate a background outline image, A first step of storing an image in a field memory; and a step of performing a calculation using a Sobel contour detection mask around each pixel with respect to the input current image when a current image photographed later from the camera is inputted after the first step A third step of calculating each pixel value corresponding to the output contour image using each pixel value of the calculated input contour image and the stored background contour image; And outputting an output contour image by using the output contour image All.

As described above, the apparatus for clearing the background in the image of the present invention calculates and stores a background contour image in advance in the input background image signal, and then calculates a current contour image of the current image to be input. Then, an output contour image is obtained using the current contour image and the previously stored background contour image, and the obtained output contour line image is displayed on the display unit.

Here, the output contour image represents the contour image of the new object appearing on the background contour image. Accordingly, the user can easily recognize the newly displayed object with respect to the background image stored in advance through the output contour image. In addition, the object extracting apparatus of the present invention can remove unnecessarily large number of contour lines from a complicated background contour image and display only the outline of a newly displayed object, thereby improving the accuracy and efficiency of image processing.

1 is a block diagram of an apparatus for subject extraction in an image according to an embodiment of the present invention.
FIGS. 2A to 2C illustrate examples of extracted contour images according to an embodiment of the present invention.
3 is a control flowchart of an apparatus for subject extraction according to an embodiment of the present invention.
4A and 4B are Sobel masks used for extracting a contour image from an image input from an image according to an embodiment of the present invention.
5 is a diagram for explaining a method of obtaining an output contour image using the extracted background contour image and current contour image according to an embodiment of the present invention.

Hereinafter, preferred embodiments of the present invention will be described in detail with reference to the accompanying drawings. It is to be noted that the same elements among the drawings are denoted by the same reference numerals whenever possible. In the following description, well-known functions or constructions are not described in detail since they would obscure the invention in unnecessary detail.

1 is a block diagram of an apparatus for subject extraction in an image according to an embodiment of the present invention. Referring to FIG. 1, an apparatus for extracting a subject in an image according to an embodiment of the present invention includes a camera 2, a preprocessing unit 4, a control unit 6, an outline detection unit 8, a field memory 10, 12).

The camera 2 focuses the optical image and picks up the incident optical image through photoelectric conversion.

The preprocessing unit (4) preprocesses the video signal output from the camera (2). According to an embodiment of the present invention, the preprocessing unit 4 performs contrast enhancement, noise removal, geometric correction and contour enhancement processing to compensate for distortion of the image signal output from the camera 2 .

The control unit 6 calculates and stores the background contour image in advance through the image input from the preprocessing unit 4, and calculates the current contour image from the input image, that is, the current image. Then, the controller 6 obtains the pixel value of the output contour image using the pixel value of the calculated current contour image and the pixel value of the previously stored background contour image, and outputs the output contour image signal corresponding to the obtained pixel value.

The contour detection unit 8 calculates an outline image from a video input from the preprocessing unit 4 under the control of the control unit 6. [ According to the embodiment of the present invention, the contour image detecting unit 8 uses a Sobel edge detection mask.

The Sobel outline detection mask (hereinafter referred to as "Sobel mask") is an algorithm for differentiating each pixel value of the corresponding image, once for each of the x-axis and the y-axis, using a second-order differential operator. The outline image detecting unit 8 calculates the outline image through the image input from the preprocessing unit 4 using the Sobel mask and outputs the calculated outline image signal.

The field memory 10 receives the outline video signal output from the outline detection unit 8 under the control of the control unit 6, and stores it as a background outline image for one field.

The display unit 12 outputs an outline signal corresponding to the output contour image obtained under the control of the control unit 6. [

According to the embodiment of the present invention, the object extracting apparatus in the image as described above calculates and stores the background contour image in advance in the input background image signal, and then calculates the input contour image for the current image to be input. Then, the subject extraction apparatus obtains an output contour image using the calculated current contour image and a previously stored background contour image, and outputs a signal corresponding to the obtained output contour image.

FIGS. 2A to 2C are examples of contour images generated according to an embodiment of the present invention. 2A through 2C, FIG. 2A illustrates an example of a background contour image calculated using a background image signal input by an object extracting apparatus according to an exemplary embodiment of the present invention. FIG. 2B illustrates an example of a current contour image calculated for a current image input by the subject extracting apparatus according to an embodiment of the present invention. 2C shows an example of an output contour image obtained by using the calculated current contour image as shown in FIG. 2B and the background contour image as shown in FIG. 2A according to an embodiment of the present invention. The output contour image shown in FIG. 2C is obtained by extracting a contour image of a newly-appearing subject 100 in the current contour image shown in FIG. 2B in contrast to the background contour image shown in FIG. 2A. The user can easily recognize the newly displayed subject 100 with respect to the previously stored background contour image through the output contour image shown in FIG. 2C.

FIG. 3 is a flowchart illustrating an apparatus for extracting a subject in an image according to an exemplary embodiment of the present invention. FIGS. 4A and 4B illustrate a Sobel mask used for extracting a contour image from an image input to an image according to an embodiment of the present invention. Fig. FIG. 5 is a diagram for explaining a method of obtaining an output contour image using a extracted background contour image and a current contour image according to an embodiment of the present invention.

1 to 5, an object extracting apparatus according to an exemplary embodiment of the present invention displays an output contour image using a background contour image for a background image input first and a contour image for a current image input thereafter Describe the process.

The control unit 6 calculates the background contour image in step 20 of FIG. 3, and stores the calculated contour image in the field memory 10. That is, the control unit 6 calculates the background contour image through the contour detecting unit 8 when the background video signal photographed from the camera 2 is input. At this time, the contour detecting unit 8 calculates each pixel value of the background contour image by performing calculation using the Sobel mask on each pixel with respect to the background image signal photographed from the camera 2 under the control of the control unit 6. That is, the contour detecting unit 8 detects the background image using the 3 × 3 Sobel mask (vertical direction) as shown in 3a and the 3 × 3 Sobel mask (horizontal direction) as shown in FIG. And calculates each pixel value of the outline image. More specifically, the contour detecting unit 8 multiplies all of the pixels of the input background image by a 3 × 3 horizontal direction Sobel mask as shown in FIG. 3A in the horizontal direction, The average of the values is obtained. Then, all the pixels of the inputted background image are multiplied by a 3 × 3 vertical direction Sobel mask as shown in FIG. 3B in the vertical direction, and an average of 8 surrounding pixels around the center pixel is obtained. Add to the calculation result. Then, the contour extracting unit 8 calculates the average of each pixel value for contour enhancement and performs threshold value processing using the calculated average value. The boundary value processing means that the pixel value is set to 1 when the pixel value of the background contour image is equal to or more than the average value of each pixel value and is set to 0 otherwise.

Equation 2

If f (x, y) > = t, g (x, y) = 1

If f (x, y) <t, g (x, y) = 0

In Equation (2), f (x, y) and g (x, y) represent the pixel values of the position (x, y) before and after the Sobel contour detection and t represents the threshold value or average value. The boundary value processing may be reversed on the basis of the boundary value in Equation (2). That is, if each pixel value of the background contour image is equal to or greater than the average value of each pixel value, the pixel value can be processed as 0, otherwise, it can be processed as 1.

The control unit 6 stores each pixel value of the background contour image calculated as described above in the field memory 10. Then, the control unit 6 receives the current video signal photographed from the camera 2 in step 22. The control unit 6 receives the current image signal photographed from the camera 2, and then proceeds to step 24 to calculate the current contour image from the received current image signal. At this time, the control unit 6 calculates the current contour image through the contour detecting unit 8. The manner in which the contour detection unit 8 detects the contour image is similar to the method of detecting the contour image described above. That is, the contour extracting section 8 multiplies all the pixels of the current image by a 3x3 horizontal direction Sobel mask as shown in FIG. 3A in the horizontal direction, and then obtains an average of eight surrounding pixel values around the middle pixel . Then, all the pixels of the input current image are multiplied by a 3 × 3 vertical direction Sobel mask as shown in FIG. 3B in the vertical direction, and an average of 8 pixel values around the center pixel is obtained. Add. Then, the contour extracting unit 8 calculates an average of each pixel value of the input current image and performs threshold value processing using the calculated average value. At this time, if the pixel value of the current image is equal to or higher than the average value, the contour extracting unit 8 sets the pixel value of the current contour image to 1, and otherwise, Here, if the pixel value of the current contour image is equal to or greater than the average value of each pixel value, the pixel value may be treated as 0, otherwise, the pixel value may be treated as 1.

The control unit 6 calculates the pixel value of the current contour image from the current image signal input from the preprocessing unit 4 as described above, and then proceeds to step 26. [ The controller 6 calculates the pixel value of the output contour using the pixel value of the current contour image calculated in step 26 and the pixel value of the previously stored background contour. That is, as shown in FIG. 4, the controller 6 calculates the pixel value of the output contour line as '0' when the pixel value of the background contour line stored in advance is '0' and the pixel value of the current contour line is '0'. If the pixel value of the background contour stored in advance is '0' and the pixel value of the current contour line is '1', the control unit 6 calculates the pixel value of the output contour line as '1'. If the pixel value of the background contour line is '1' and the pixel value of the current contour line is '0', the controller 6 calculates the pixel value of the output contour line as '0', and if the pixel value of the background contour line is '1' , The pixel value of the output contour line is calculated as '1' when the pixel value of the current contour line is '1'. When the pixel value of the output contour image is obtained, the control unit 8 displays the output angle image corresponding to the pixel value obtained through the display unit 12 in step 28. At this time, the output contour image displayed on the display unit 12 is preferably the contour image 100 of the object newly displayed in the background contour image.

While the present invention has been described in connection with what is presently considered to be practical exemplary embodiments, it is to be understood that the invention is not limited to the disclosed embodiments. Therefore, the scope of the present invention should not be limited by the described embodiments but should be determined by the equivalents of the claims and the claims.

2:
4:
6: contour image calculating section
8:
10: Field memory
12:

Claims (7)

An apparatus for erasing a background in an image,
A contour line detecting unit for calculating contour line images from the image photographed by the camera; and a contour detecting unit for receiving the contour line image calculated from the contour detecting unit, A background memory for storing the generated background contour image in the field memory when the background image photographed first from the camera is inputted, and the background contour image is calculated in the input background image, the calculated background contour image is stored in the field memory, And a controller for calculating a current contour line image from the input current image when the current image is input and outputting an output contour line image using the calculated current contour line image and the stored background contour line image, / RTI &gt;
The method according to claim 1,
Further comprising: a preprocessor for performing preprocessing of the video signal input from the camera and outputting the preprocessed video signal.
The method according to claim 1,
And a display unit for displaying an output contour image output from the control unit.
3. The method of claim 2,
Wherein the preprocessor performs at least one of brightness adjustment, noise removal, geometric correction, and contour enhancement processing on the image signal output from the camera.
The method according to claim 1,
Wherein the contour detection unit calculates a contour line image using a Sobel contour detection mask.
6. The method of claim 5,
Wherein the contour detection unit calculates a contour line image for the input image by performing calculation using a Sobel contour detection mask for the input image when a captured image is input from a camera, .
6. The method of claim 5,
Wherein the contour detection unit calculates each pixel value of the input image by performing calculation using a Sobel contour detection mask around each pixel input to the input image when a captured image is input from the camera, Calculating an average value of the pixel values and outputting the output contour image by setting the value of the corresponding pixel to 1 if the value of the pixel is equal to or greater than the average value and setting the value of the corresponding pixel to 0 if the value of the corresponding pixel is less than the average value; A device for erasing a background from a feature image.

Images