KR100813936B1 - Method for extracting subject and image synthesizing in moving picture - Google Patents

Abstract

The present invention relates to a method of extracting a dynamic subject of a video and a service for synthesizing a video. The present invention detects and extracts a moving user's subject from a user's video transmitted through a web, and transparently processes a surrounding background image to generate a corresponding user's single image. By synthesizing various background images in a single video and providing them to the user through the web, users can compose various background images in their own images and give users interesting and fun at the same time through the dynamic user appearance in interesting background images. It is characterized by being.

Video, transparent

Description

Dynamic subject extraction and image synthesis service method of video {METHOD FOR EXTRACTING SUBJECT AND IMAGE SYNTHESIZING IN MOVING PICTURE}

1 is an exemplary view illustrating a system to which a method for extracting dynamic subjects and a video synthesis service of a video of the present invention is applied.

Figure 2 is an exemplary view for explaining a dynamic subject extraction process of a video according to the present invention.

3 is an exemplary diagram for explaining an example of designating a region of interest within an image according to the present invention;

Figure 4 is a flow diagram illustrating a dynamic subject extraction and image synthesis service method of a video according to an embodiment of the present invention.

The present invention relates to a dynamic subject extraction and image synthesis service method of a video, and more particularly, to detect and extract a moving user subject from a user's video transmitted through a web, and to process the surrounding background image by transparently processing the user's single image. By creating and synthesizing various background images on the user's sole video and providing them to the user through the web, users can compose various background images on their own images to make the user interesting and fun through the dynamic user image in the interesting background image. The present invention relates to a dynamic subject extraction and a video synthesis service method capable of simultaneously giving an image.

In general, image processing technology, which has recently been spotlighted in various fields, refers to a series of processes that receive image data from an image device (eg, a digital camera, a scanner, a PC camera) and process the image data.

Typically, such image processing techniques include image synthesis techniques that are widely used in fields such as computer art, advertising, and entertainment by combining or transforming real image images with images of artificial images, and objects appearing in image images. Image Analysis technology, which refers to a series of tasks to identify the image, and Image Pattern Recognition technology, which refers to the task of identifying and recognizing objects such as a character, a human fingerprint, and a face given as an image image. have.

An object of the present invention is to detect and extract a moving user's subject from a user's video transmitted through the web, and to process the surrounding background image to generate the corresponding user's single image, and to synthesize the various background images to the user's single image. By providing to the user through the user to provide a dynamic subject extraction and image synthesis service method for the video that allows the user to synthesize a variety of background images to his image.

Another object of the present invention is to provide a method for extracting dynamic subjects and an image synthesis service of a moving image which enables high-speed transparency processing and prevents computational overload in extracting dynamic subjects performed through the web.

According to an aspect of the present invention, a dynamic subject extraction and image synthesis service method according to an aspect of the present invention includes (a) an initial user exclusion background from which a dynamic subject is excluded from a user computer connected to an operation management server through an information communication network; Receiving an image; (b) designating, by the operations management server, a plurality of ROIs for the received initial user-excluded background image, and storing position coordinate values and color values for the ROIs; (c) receiving, by the operations management server, image frames of a subject image including a dynamic subject while setting a background image of the initial non-user background from the user computer; (d) the operation management server extracts a location coordinate value and a color value in the image of the regions of interest defined for the background image excluding the initial user, and calculates the coordinate values and the color values of the regions of interest. Determining whether the color value is changed by comparing the color values of the ROIs by substituting the image frames of the image frames; (e) the operation management server classifying the region of interest for which there is no change in the color value based on the existing initial user-excluded background image with respect to the image frame and the region of interest for which the change in the existing color value occurs; (f) The operation management server confirms the region of interest that does not change color value as the color unchanged area compared to the initial user excluded background image, and clears the color unchanged area by deleting the color unchanged area from the corresponding image frame. Processing to extract a subject-only image; And (g) generating, by the operation management server, a user-only image layer by combining the subject-only images loaded according to the respective image image frames sequentially received in chronological order. Characterized in that it comprises a.

Preferably, in the step (b), the ROIs are designated as a grid pattern spaced apart from each other by a predetermined interval in the entire initial user excluded background image.

More preferably, in step (b), when designating the regions of interest, the regions of interest are designated as central regions of interest in a grid pattern spaced apart from each other at a predetermined interval and at the same time for each of the central regions of interest. Designate a predetermined number of neighboring neighboring interest zones, quantify the difference in color values between the central interest zone and the neighboring interest zones, sum the difference of the respective color values between each central interest zone and its surrounding interest zones, and store the corresponding sum. Making; It characterized in that it further comprises.

More preferably, after step (d), the total sum of respective color value differences between the pre-stored central region of interest and its surrounding regions of interest for the initial user-excluded background image and the correspondence in the corresponding received image frame. Comparing the total sum of the respective color value differences between the central ROI and the neighboring ROIs, and determining that the central ROI is occupied by the subject if there is a difference in the sum of the sums exceeding a predetermined reference value; It characterized in that it further comprises.

More preferably, the operation management server to set the object detection sensitivity for the region of interest to determine whether the dynamic subject in the frame image of the video captured by the user computer; It characterized in that it further comprises.

More preferably, the region of interest is one pixel or a predetermined number of pixel aggregates in the image.

More preferably, after the step (g), when (h) the request signal for adding the background image is received from the user computer, the operation management server synthesizes and provides a predetermined background image to the generated user single image layer. Making; It characterized in that it further comprises.

More preferably, (i) when the additional request signal of the background sound is received from the user computer, the operation management server synthesizes and provides a predetermined background sound to the generated user-only image layer; It characterized in that it further comprises.

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

Hereinafter, the present invention detects and extracts a moving user's subject from a user's video transmitted through the web, transparently processes the surrounding background image to generate a corresponding user's single image, and synthesizes various background images on the user's single image through the web. By providing the user with a variety of background images to the user's own video to create a dynamic subject extraction and video synthesis service method that can give the user interesting and fun at the same time through the dynamic user image in the interesting background image Although it will be described as a preferred embodiment, the technical idea of the present invention is not limited to this, but can be variously modified and implemented by those skilled in the art.

1 is an exemplary view illustrating a system to which a method for extracting dynamic subjects and a video synthesis service of a video of the present invention is applied.

The system to which the dynamic subject extraction and image synthesis service method of a video of the present invention is applied is a user computer 100 for capturing an image of a user and transmitting it to the operation management server 200 through a web and the user computer 100 through a web. Recognizing and deleting the background other than the user having dynamic movement in the video image frame of the user received from the transparent) and the operation management server 200 to synthesize a variety of background images to the image of the user alone the background is transparent It is composed.

Referring to Figure 1 in more detail with respect to the dynamic subject extraction and video synthesis service method of a video according to an embodiment of the present invention,

The user computer 100 is connected to the operation management server 200 through an information communication network including a wired Internet network of TCP / IP protocol and a wireless Internet network such as WAP protocol, and is operated by the operation management server 200. By executing the image synthesis service to shoot a video of the user of the user computer 100 and transmits to the operation management server 200, the user computer 100 is a computer device that can communicate web content to display A mobile phone, a personal digital assistant (PDA) and an IMT2000 and a landline phone capable of wireless internet communication as well as a personal computer such as a desktop computer, a notebook computer, and the like, not shown in the drawings. Device, at least Pentium-class CPU, data storage, 64Mb memo And it includes a device, Windows 98 above the operating system.

That is, the user computer 100 photographs the video image of the user through the corresponding user computer 100 and transmits the image to the operation management server 200 through the information communication network in real time. Module 110.

In this case, the user computer 100 is preferably a webcam (WebCam) (or a PC camera or a USB camera) that can be immediately employed without a separate hardware technology in a general PC, but the captured image is digitized and the computer through a predetermined interface. It does not matter if it is a general digital camera to transmit to.

Meanwhile, the operation management server 200 receives a video image frame of the user transmitted from the user computer 100 and extracts the user's single image by transparently processing a background other than the user from the video image to use the user's single image. As shown in FIG. 1, the main controller 210, the data storage device 220, the external interface 230, and the sensor setting unit 240 are synthesized and provided to the corresponding user computer 100. ), An image processor 250, and an image synthesizer 260.

The data storage device 220 includes a user image DB for storing a user-only image whose background is transparent from an original image transmitted from users, and a background image DB for storing various background images for synthesis with the user-only image; And a background sound DB for storing various background sounds for synthesizing with the user only image, and a synthesized image DB for storing a synthesized image obtained by synthesizing a background image and a background sound in the user only image for service to a user.

In this case, the background image refers to a background image (eg, a music video, a beach image, a recreational forest image, a studio photographing field image, etc.) synthesized as a back background with respect to a user sole image processed as a layer, and a front background for a user sole image. It includes a front background image (for example, a frame image, a border image) to be synthesized, and an accessory image (for example, a bird, a flower, a butterfly, a bee, etc.) that is additionally synthesized as a front or back background of a user alone image.

The accessory image is processed as a separate image layer to enable dynamic movement in the image.

In addition, the background sound refers to various background sounds (for example, waves, train sounds, music, etc.) that can be synthesized.

The external interface 230 configures a connection interface with the user computer 100 connected through the information communication network, and receives a video image frame transmitted in real time from the user computer 100.

The sensor setting unit 240 stores the initial user-excluded background image photographed and received from the user computer 100, and designates a plurality of ROIs in a plane from the initial user-excluded background image to determine the position coordinate values and The color value is stored and managed, and includes an initial background storage unit 241, a region of interest designation unit 242, and a region of interest storage unit 243.

The initial background storage unit 241 stores the initial user excluding background image received from the user computer 100 through the external interface 210.

In this case, the initial user-excluded background image refers to a static background image that does not include a user who is a dynamic subject as illustrated in FIG. 2A.

The region of interest designator 242 may designate a plurality of regions of interest for checking the movement when the user dynamically moves in the same background image in the initial user excluded background image.

In this case, as shown in FIG. 2B, the region of interest is a position of a sensor for detecting a dynamic movement of a user in a background image excluding the initial user, and this one region of interest is designated as one pixel or a predetermined number of pixels. Can be.

The ROI storage unit 243 stores position coordinates and color values in the background images of the ROIs designated by the ROI designator 242 for the initial user-excluded background image.

Here, the region of interest is an index for recognizing the movement of the dynamic subject through the change of the color value of the region of interest in the image image frame to which the movement of the user, which is the dynamic subject, is added, based on the background image. This area of interest, however, should be minimized for smooth computation of real-time image processing.

Therefore, the region of interest is designated as a lattice pattern in the background image excluding the entire initial user, as shown in FIG. 3A, so that only a predetermined number of regions of interest are monitored instead of all pixels, and the color of the region of interest is changed. You can see the movement.

However, when specifying the grid-shaped region of interest, the number of regions of interest in the grid-patterned name is limited in order to prevent the overload of computation for real-time image processing.

Accordingly, the region of interest specifies a plurality of central regions of interest in a lattice pattern within the entire initial user-excluded background image as shown in FIG. 3b and also includes eight adjacent regions in eight directions for each of the central regions of interest. You can specify a region of interest around you.

In this case, when the region of interest is designated, the color values of the corresponding center region of interest and the eight surrounding regions of interest are respectively calculated, and the eight color differences between the center region of interest and the eight surrounding regions of interest are calculated to obtain these eight differences. The sum is stored and stored in the ROI storage unit 24. In the future, when the user's movement is checked, the center ROI is occupied by the dynamic subject when the center ROI is over the predetermined threshold by comparing the sum of the color difference between the corresponding center ROI and the neighboring ROI and the previously stored color difference. It is decided that it is.

That is, in order to more accurately determine whether a central region of interest is occupied by a dynamic subject, the color value of the central region of interest is primarily compared with the color value in the pre-stored initial user-excluded background image. Determines whether the subject is occupied, and secondarily compares the sum of the difference in color values of the corresponding center interest area and the surrounding interest area with the sum of the difference in color values in the pre-stored background image except for the initial user. It will determine whether or not it is occupied.

In this way, it is possible to more accurately determine whether the subject occupies even if the number of designated areas of interest, that is, the detection zone, is small.

Here, the region of interest may be designated as four neighboring regions of interest in the four-direction direction with respect to each of the central regions of interest in the entire initial user excluded background image, as shown in FIG. 3C.

In this case, the color value means an RGB value of each pixel.

The process of determining whether the subject occupies the region of interest may be better understood through the following description.

Meanwhile, the image processor 250 receives and stores image frames of a video including a user captured and received from the user computer 100 in real time, and changes the color of the ROI for each image frame, that is, RGB. The dynamic subject is recognized in the image by changing the value, and the image of the dynamic subject is extracted for each frame. The image frame storage unit 251, the ROI comparison unit 252, and the image region classification unit 253 And a subject extracting unit 254.

The image frame storage unit 251 sequentially receives and stores image frames of an image transmitted from the user computer 100.

In addition, the ROI comparison unit 252 obtains the location coordinate value and the color value in the image of the ROI predetermined for the background image excluding the initial user from the ROI storage unit 243 and the image frame storage unit 251. By comparing the color values of the respective ROIs by substituting the image frames stored in), it is determined whether the color values of the ROIs that determine whether the ROIs are existing backgrounds or dynamic objects are changed.

In addition, the region of interest comparison unit 252 further extracts the sum of the color value differences between the central region of interest and each of the surrounding regions of interest and the previously stored initial user exclusion background image to extract the dynamic subject more accurately as described above. It is also possible to determine whether the subject occupies the center region of interest by comparing the sum of the color value differences between the central region of interest and each surrounding region of interest.

In general, an error may occur due to a change in light rays, a shadow, or automatic color adjustment of a camera in a real image. Therefore, a reference for the degree of change in color values for determining whether a subject is a target is required. do.

The image region classifying unit 253 classifies each region by classifying the region of interest having no change in the existing color value and the region of interest in which the change has occurred in the existing color value according to the determination result of the region of interest comparison unit 252. Confirm.

At this time, for the ROIs that do not change from the existing background color value, that is, the ROIs of the background, the area up to the other ROI within the range of RGB color within a predetermined threshold based on its own color value for each ROI. Expand to determine the background areas for each color. Each of these background areas is then summed into the entire transparent area.

The subject extracting unit 254 erases the entire background region classified by the image region classifying unit 253 to determine the region, thereby extracting only the dynamic subject image by making the background region transparent.

By repeatedly loading this process for each received image frame, a single image layer of a user, which is a dynamic subject, is generated.

The image synthesizing unit 260 synthesizes a separate image by synthesizing various backgrounds (front background, back background, accessory images) stored in the background image DB according to a user's request for the user's single image layer.

The overall flow of the present invention will be described with reference to FIG. 4 as follows.

First, the main control unit 210 of the operation management server 200 sets a setting for a subject detection sensitivity of a region of interest for determining whether a user is a dynamic subject or a user in a frame image of a video photographed by the user computer 100. Perform (S11).

In this case, the object detection sensitivity is that each region of interest by comparing the region of interest of the frame image stored in the image frame storage unit 252 with the region of interest of the background image excluding the initial user is stored previously. In this embodiment, the sensitivity of the subject detection may be a pixel value, that is, an RGB value. Therefore, by setting the color change allowable width of the pixel value of the region of interest, if a change in the pixel value exceeding the allowable width occurs, the region of interest comparison unit 252 recognizes that the region of interest of the subject occupies the region of interest.

Subsequently, the main controller 210 requests the user computer 100 connected through the information communication network to photograph the background image excluding the initial user excluding the user as the dynamic subject, and excludes the initial user photographed from the user computer 100. The background image is received and stored in the initial background storage unit 241 (S12).

The background image excluding the initial user means an image photographed against a background that does not include a user, which is a moving dynamic subject, as shown in FIG. 2A. Of course it moves.

The main control unit 210 extracts an initial user excluding background image stored in the initial background storage unit 241 through the ROI designation unit 242, and designates a plurality of ROIs for the corresponding initial user excluded background image; 2B), the location coordinate values and the color values of the ROIs are registered in the ROI storage unit 243 (S13).

In this case, the ROIs may be designated as a grid pattern in the entire initial user excluded background image, as described above (see FIG. 3A).

In addition, in order to increase the accuracy of object detection, a plurality of central points of interest are designated in a grid pattern within the entire initial user-excluded background image, and additionally, eight adjacent points of interest are adjacent to each of the center points of interest in eight directions. It may also be specified (see Figure 3b).

Thereafter, the main controller 210 requests the connected user computer 100 to capture an image including the corresponding user in the background except for the initial user, that is, an image including a dynamic subject in the background except for the initial user (S14). .

In this case, it is preferable that a background having a minimum complexity is selected in order to prevent an overload of a system and a time delay of an operation.

Subsequently, the main controller 210 receives image frames (see FIG. 2C) of images captured and transmitted from the user computer 100 and stores them in the image frame storage unit 251 (S15). The region comparison unit 252 takes the position coordinate values and the color values in the image of the region of interest defined for the background image excluding the initial user, and substitutes the image frames stored in the image storage unit 241 for each interest of interest. The color values of the zones are compared to determine whether each of the zones of interest are existing backgrounds or dynamic subjects (S16).

The result of the determination of whether or not the dynamic object of each of the ROIs is shown in FIG. 2D.

In this case, in step S16, in order to increase the accuracy of object detection, the central interest zone and each peripheral interest zone in the pre-stored initial user-excluded background image and the sum of the color value differences between the central interest zone and each peripheral interest zone are stored. The step of determining whether the subject occupies the center ROI may be further performed by comparing the sum of the color value differences between the two.

In addition, the main control unit 210 is configured to generate a region of interest and an existing color value that are not changed in the color value based on the existing initial user-excluded background image through the image region classifier 253 according to the determination result of step S16. The region of interest in which the change has occurred is classified, and each region is determined as a color change region and a color non-change region as shown in FIG. 2E (S17).

When the color change area and the color non-change area are determined, for a region of interest in which no change has occurred in the existing background color value, that is, a region of interest in the background, a predetermined threshold value is determined based on its own color value for each region of interest. The background areas for each color are determined by extending the area to other regions of interest within the RGB color range.

Subsequently, the main control unit 210 deletes the color non-change region from the corresponding captured image frame through the subject extracting unit 254 according to the classification of the color change region and the color non-change region of step S17. The non-variable region is made transparent so that only a single image of a subject, that is, a user, is selected and extracted as shown in FIG. 2F (S18).

As described above, steps S15 to S18 are performed on the respective image image frames sequentially transmitted from the user computer 100, so that the user's single images are sequentially loaded. By combining to generate a user-only image layer is stored in the user image DB of the data storage device 220 (S19).

Subsequently, the main controller 210 receives a user theme addition request signal for requesting the addition of a background image, the addition of an accessory image, or the addition of a background sound to the user-only image layer generated through the above-described process from the corresponding user computer 100. It is determined whether the received (S20), if there is a request for the theme addition of the user proceeds to the following various theme addition process.

When a request signal for adding a background image is received from the user computer 100, the main control unit 210 of the operation management server 200 is a background image list or user synthesized as a background for the user's single image layer. The front background image list synthesized as the front background is transmitted to the single image layer to request the user to select an additional theme image (S21).

Here, the back background image refers to a back background image (eg, a music video, a beach image, a recreational forest image, a studio photographing field image, etc.) synthesized as a back background with respect to a user-only image processed as a layer. Means a front background image (for example, a frame image and a border image) synthesized with the front background with respect to the user's single image.

Subsequently, the main control unit 210 sets the background image selected by the user to the underview having a high display order with respect to the generated user single image layer through the image synthesizing unit 260 (see FIG. 2G), or the user. Synthesizes the background image selected from the user's sole image layer with the low-order overview of the generated user's single image layer through the image synthesis unit 260 (see FIG. 2H), thereby synthesizing the user's single image layer and the background image. It is made (S22).

At this time, it is also possible that the back background image and the front background image may be simultaneously synthesized with respect to the user's single image layer.

On the other hand, when the additional request signal of the accessory image is received from the user computer 100, the main control unit 210 of the operation management server 200 is an accessory that is synthesized with the front background or the back background for the user alone image layer The video list is transmitted to request the user to select an additional theme video (S23).

Here, the accessory image includes an accessory image (for example, a bird, a flower, a butterfly, a bee, etc.) that is additionally synthesized as the front or back background of the user alone image.

The accessory image is processed as a separate image layer to enable dynamic movement in the image.

Subsequently, the main control unit 210 displays the accessory image selected by the user through the image synthesizing unit 260 with an underview having a high display rank with respect to the generated user sole image layer or with a low display rank with respect to the user sole image layer. By synthesizing by setting (see FIG. 2I), the user's single image layer and the accessory image are synthesized (S24).

At this time, it is also possible that the background image, the front background image, and the accessory image may be simultaneously synthesized with respect to the user's single image layer.

On the other hand, when the additional request signal of the background sound theme is received from the user computer 100, the main control unit 210 of the operation management server 200 transmits the background sound list synthesized as the background sound for the user's sole image layer The user is asked to select an additional theme background sound (S25).

Here, the background sound refers to various background sounds (eg, waves, train sounds, music, etc.) that can be synthesized.

Subsequently, the main controller 210 synthesizes the background sound selected from the user as the playback background sound of the generated user single image layer through the image synthesizing unit 260, thereby synthesizing the background sound with the corresponding user single image layer (S26). ).

At this time, it is also possible that the background image, the front background image, the accessory image, and the background sound can be simultaneously synthesized with respect to the user single image layer.

As described above, the main control unit 210 of the operation management server 200 converts the resultant image converted into a composite multilayer image by adding a theme to the user's single image layer by the user's selection. Will be provided as (S27).

The present invention described above is capable of various substitutions, modifications, and changes without departing from the spirit of the present invention for those skilled in the art to which the present invention pertains. It is not limited to drawing.

As described above, in the dynamic subject extraction and image synthesis service method of the present invention, a moving user's subject is detected and extracted from the user's video transmitted through the web, and the surrounding background image is transparently processed so that the user alone. By creating images and synthesizing various background images on the user's sole image and providing them to the user through the web, users can compose various background images on their own images to interest the user through the dynamic user image in the interesting background image. And fun at the same time has the effect.

In addition, when designating a region of interest that detects a color change in dynamic subject extraction through a web, the region of interest is arranged in a grid shape with a predetermined pixel interval in the image instead of the entire image pixel, or a background image. Designate multiple central points of interest in a grid pattern within the grid and additionally specify eight adjacent points of interest in eight directions for each central point of interest, enabling high-speed transparent processing in real time and preventing computational overload. In addition, there is an effect that can increase the accuracy of the overall object detection.

Claims (8)

(a) receiving, by the operation management server, an initial user-excluded background image excluding a dynamic subject from a user computer connected through an information communication network; (b) designating, by the operations management server, a plurality of ROIs for the received initial user-excluded background image, and storing position coordinate values and color values for the ROIs; (c) receiving, by the operations management server, image frames of a subject image including a dynamic subject while setting a background image of the initial non-user background from the user computer; (d) the operation management server extracts a location coordinate value and a color value in the image of the regions of interest defined for the background image excluding the initial user, and calculates the coordinate values and the color values of the regions of interest. Determining whether the color value is changed by comparing the color values of the ROIs by substituting the image frames of the image frames; (e) the operation management server classifying the region of interest for which there is no change in the color value based on the existing initial user-excluded background image with respect to the image frame and the region of interest for which the change in the existing color value occurs; (f) The operation management server confirms the region of interest having no color value change as the color unchanged area compared to the background image excluding the initial user, and clears the color unchanged area by deleting the color unchanged area from the corresponding image frame. Processing to extract a subject-only image; And (g) generating, by the operation management server, a user-only image layer by combining the subject-only images loaded according to the respective image image frames sequentially received in chronological order; Dynamic subject extraction and image synthesis service method of a video comprising a. The method of claim 1, In step (b), And the ROIs are designated in a grid pattern spaced apart from each other within a background image of the entire initial user except the initial user. The method of claim 2, In step (b), In designating the regions of interest, the regions of interest are designated as central regions of interest in a lattice pattern spaced apart from each other, and a predetermined number of neighboring regions of interest for each central region of interest are designated and centered. Quantifying the difference between the color values between the ROI and the surrounding ROI, and summing the difference of the respective color values between each central ROI and the surrounding ROI, and storing the sum value; Dynamic subject extraction and image synthesis service method of the video further comprising a. The method of claim 3, wherein After step (d), The total sum of the difference of each color value between the pre-stored center interest zone and its surrounding interest zones for the initial user-excluded background image, and the respective center interest zone and its surrounding interest zones in the corresponding received image frame. Comparing the total sum of the color value differences and determining that the corresponding central ROI is occupied by the subject when there is a difference in the sum values exceeding a predetermined reference value; Dynamic subject extraction and image synthesis service method of the video further comprising a. The method of claim 1, Before step (a) above, Setting, by the operations management server, a subject detection sensitivity of a region of interest for determining whether a dynamic subject is in a frame image of a video photographed by the user computer; Dynamic subject extraction and image synthesis service method of the video further comprising a. The method according to any one of claims 1 to 5, And the region of interest is one pixel or a predetermined number of pixels in the image. The method of claim 1, After step (g), (h) when the request signal for adding a background image is received from a user computer, the operation management server synthesizing and providing a predetermined background image to the generated user only image layer; Dynamic subject extraction and image synthesis service method of the video further comprising a. The method of claim 7, wherein (i) when the request signal for adding a background sound is received from a user computer, the operation management server synthesizing and providing a predetermined background sound to the generated user single image layer; Dynamic subject extraction and image synthesis service method of the video further comprising a.

Images