WO2017209432A1

WO2017209432A1 - Randomly-edited compressed video content provision system and provision method

Info

Publication number: WO2017209432A1
Application number: PCT/KR2017/005356
Authority: WO
Inventors: 이규영; 천솔지
Original assignee: (주)잼투고
Priority date: 2016-05-31
Filing date: 2017-05-23
Publication date: 2017-12-07
Also published as: KR101843017B1; KR20170135299A

Abstract

The present invention relates to a randomly-edited compressed video provision system and provision method. The randomly-edited compressed video content provision system, according to the present invention, comprises: a video content storage unit for storing compressed video content comprising a first GOP to an N^th GOP (N is an integer greater than or equal to 2), each GOP comprising a first non-reference frame which does not refer to information on a preceding frame, and multiple reference frames which refer to information on a preceding or succeeding frame; and a video content provision unit for providing the compressed video content in GOP units to a user terminal via a communication network, wherein at least some GOPs are provided to the user terminal in order, and a k^th GOP (k is an integer greater than or equal to 2 and less than or equal to N-1) is provided to the user terminal in a random order preceding a (k-1)^th GOP or succeeding a (k+1)^th GOP.

Description

Arbitrary editing compressed video contents provision system and method

The present invention relates to a system and a method for providing a random edited compressed video content, and more particularly, to a random edited compressed video content providing system which can be provided to a user by randomly editing in a GOP unit starting with a best-order non-reference frame during random editing. And a method of providing the same.

Transmission of various video contents through a communication network has become common. The uncompressed video content file, which contains a complete image of every frame, has a problem of increasing network resources and data transmission costs required for video content transmission due to excessive file size. Accordingly, video storage format standardization organizations such as the Moving Picture Experts Group (MPEG) are proposing various compressed video content storage standards.

According to the video compression standards such as MPEG-2, MPEG-4, ITU-T H.263, ITU-T H.264 / MPEG-4, Part 10, Advanced Video Codec (AVC), as shown in FIG. The content is divided into, for example, GOP (Group of Picture) units having a length of about 0.5 seconds, and one GOP is changed in an intra frame, such as an I frame having complete image information, and a frame before or after it. It consists of a reference frame (inter frame) such as a B frame or a P frame, which stores only information and has incomplete image information, but greatly reduces the amount of data. The video compression technique contributes to the easy distribution of video contents through the communication network by reducing the amount of data of the video contents by the GOP structure composed of non-reference frames and reference frames. (Reference, Scalable Parallel Programming Applied to H.264 / AVC Decoding, pp. 5-15)

On the other hand, UGC (User Generated Contents) produced by non-professional users is often composed of a series of monotonous images due to limitations of general users' shooting technology, shooting equipment, editing technology and editing equipment. There was a hard limit. In order to solve this problem, the monotony of video content can be improved by editing a single video or a plurality of videos through an editing technique. Korean Patent Application No. 2016-0034287 and Korean Patent Application No. 2016-0048882 filed by the present applicant provide a video content providing system that can enhance the unexpectedness of a video by automatically editing one video or a plurality of videos. It starts.

On the other hand, in the case of compressed video contents other than uncompressed video contents, if the starting point to be randomly edited is an unreferenced frame, image quality deterioration occurs due to the loss of reference data at the beginning of the randomly edited section due to the absence of a preceding reference frame. There is a problem.

Qualcomm U.S. Patent No. 9,319,448, "Trick modes for network streaming of coded multimedia data," refers to the location of a reference frame as random information (Random Access Point) for random access to video content files. By storing, the user terminal provides an effect of easily accessing a non-reference frame when searching a section of a video content file. However, according to the prior art, there is a problem in that it is not possible to automatically provide video content that is automatically edited only to provide the user with access to any point of the video content.

Ericsson's U.S. Patent No. 8,340,113 “Method and arrangement for improved media session management” indicates that the user terminal starts from the video content when the transmission mode is changed when a request is made to change the transmission mode from broadcast to unicast. By requesting a point, it provides an effect of viewing seamless video content despite a change in broadcast mode. However, according to the prior art, the user cannot watch the video content in frame order, so that the user cannot automatically receive the edited video content.

[Patent Documents]

(Patent Document 1) US Patent No. 9,319,448

(Patent Document 2) US Patent No. 8,340,113

[Non-Patent Documents]

(Non-Patent Document 1) Juurlink et al., “Scalable Parallel Programming Applied to H.264 / AVC Decoding”, 2012, pp. 5-11

The present invention is to solve the above problems, according to the random edited compressed video content providing system and method according to the present invention, when the video content providing system stores the compressed video content and provides the video content to the user, By providing the video content whose video frames are edited in GOP units according to the order of, the user terminal plays back the edited GOP edited at an arbitrary point from the best non-reference frame that does not require the video information of the preceding frame. It aims at preventing the fall of image quality.

According to the system and method for providing a random edited compressed video content according to an embodiment of the present invention, by detecting a scene change point in the video content and editing the non-reference frame detected by the scene change to an arbitrary point, the unexpected of the edited video content Its purpose is to maintain the context of the original video content while improving gender.

According to another exemplary embodiment of the present invention, a system and a method for providing an arbitrarily compressed video content may include: a GOP including an IDR frame (Instantaneous Decoder Refresh Frame) that does not affect the image quality of a preceding frame as a best-order non-reference frame; By arbitrary editing, another object of the present invention is to prevent deterioration of the quality of a subsequent frame in a preceding frame despite the gap of a GOP edited at an arbitrary point.

According to another embodiment of the present invention and a method for providing a random edited compressed video content, a video section that is too short in the edited video content is defined by randomly arranging consecutive GOPs instead of a single GOP as an edited GOP array. Another purpose is to prevent them from being randomly distributed and to maintain the context of the original video content.

According to another embodiment of the present invention and a system for providing a random edited compressed video content, a GOP including the same GOP consecutively is defined as an edited GOP array, so that automatic random editing is performed on a per-object basis. Another purpose is to maintain the context of the original content.

Finally, according to another system and method for providing a random edited compressed video content according to another embodiment of the present invention, by defining a GOP array containing the same person continuously as an edited GOP array, the original through automatic random editing in the character unit Another purpose is to be able to maintain the context of the content.

In order to achieve the above object, the random edited compressed video content providing system according to the present invention includes a plurality of reference frames that refer to information of a best-order non-reference frame that does not refer to information of a preceding frame and information of a preceding or following frame. A video content storage unit for storing compressed video content including first to Nth GOPs each configured to include N (an integer of 2 or more); And providing the compressed video content to a user terminal in units of GOPs through a communication network, wherein at least some of the GOPs are provided to the user terminal in order, and k th GOP (k is an integer of 2 or more and N-1 or less). And a video content providing unit provided to the user terminal in any order preceding the k-1 GOP or following the k + 1 GOP.

In the arbitrary edited compressed video content providing system according to an embodiment of the present invention, the best-order non-reference frame is an I frame (Infra Frame), and the reference frame is a P frame (Predicted Frame) and / or B frame (Bidirectional Frame) It is characterized by the).

In the arbitrary edited compressed video content providing system according to an embodiment of the present invention, the video content providing unit may include a GOP including an I frame located at a scene change point detected through scene change detection as the best-order non-reference frame. The kth GOP is selected.

In the arbitrary edited compressed video content providing system according to an embodiment of the present invention, the video content providing unit selects a GOP including an Instantaneous Decoder Refresh (IDR) frame as the best non-reference frame as the k-th GOP. It features.

In the arbitrary edited compressed video content providing system according to an embodiment of the present invention, the video content providing unit may include the edited GOP array defined as the kth GOP to the k + m GOP (m is an integer of 10 or more). and providing the user terminal in any order to precede the k-1 GOP or follow the k + m + 1 GOP.

The random edited compressed video content providing system according to an exemplary embodiment of the present invention further includes a video object analyzer configured to analyze the VOPs (Video Object Planes) of the best non-reference frame and the image of the reference frame, and provide the video content. The unit is characterized by defining the k-th GOP to the k + m GOP including the same VOP consecutively as the edited GOP array.

The random edited compressed video content providing system according to an exemplary embodiment of the present invention may further include a face recognizing unit that recognizes a face of a person included in an image of the best non-reference frame and the reference frame and associates it with a face identifier. The video content providing unit may define the kth GOP to the k + m GOPs associated with the same face identifier consecutively as the edited GOP array.

The random edited compressed video content providing method according to the present invention includes a plurality of reference frames that refer to information of a best-order non-reference frame that does not refer to information of a preceding frame and information of a preceding or following frame in a video content providing system. A first step of storing the compressed video content composed of the first to Nth GOPs (N is an integer of 2 or more), respectively; A second step of the video content providing system receiving a random edit request for requesting arbitrary edit of the compressed video content from a user terminal through a communication network; And the video content providing system to provide the compressed video content to the user terminal in units of GOPs through a communication network, wherein at least some of the GOPs are provided to the user terminal in order, and the kth GOP (k is equal to or greater than 2). A third step of providing the user terminal in any order preceding the k-1 GOP or following the k + 1 GOP.

In the random edited compressed video content providing method according to an embodiment of the present invention, in the first step, the video content providing system includes: the best non-reference frame is an I frame (Infra Frame), and the reference frame is P; And store the compressed video content as a frame and / or a bidirectional frame.

In the random edited compressed video content providing method according to an embodiment of the present invention, the third step may include: deactivating the I frame located at the scene change point detected by the scene change detection by the video content providing system; The GOP included in the frame is selected as the k-th GOP.

In the random edited compressed video content providing method according to an embodiment of the present invention, the third step may include the GOP including the Instantaneous Decoder Refresh (IDR) frame as the best non-reference frame. The kth GOP is selected.

In the random editing compressed video content providing method according to an embodiment of the present invention, in the third step, the video content providing system edits the k th GOP to the k + m GOP (m is an integer of 10 or more). A GOP array is defined, and the edited GOP array is provided to the user terminal in any order to precede the k-1 GOP or to follow the k + m + 1 GOP.

In the random edited compressed video content providing method according to an embodiment of the present invention, the first step is the video content providing system, the VOP (Video Object Planes) of the image of the best-order non-reference frame and the reference frame And a second sub step of analyzing, wherein the video content system defines, as the edited GOP array, the k th GOPs to the k + m GOPs including the same VOP consecutively. It is characterized by.

In the random edited compressed video content providing method according to an embodiment of the present invention, in the first step, the video content providing system recognizes the face of the person included in the image of the best-order non-reference frame and the reference frame. And a second sub-step of associating with a face identifier, wherein the second step includes: the editing GOP of the k th GOP to the k + m GOPs in which the same face identifier is continuously connected. It is characterized by defining as an array.

According to the above-described configuration of the present invention, according to the system and method for providing arbitrary edited compressed video content according to the present invention, the video content providing system stores the compressed video content and provides the video content to the user in any order. Accordingly, by providing video content in which video frames are edited in units of GOPs, the user terminal plays back edited GOPs edited at random points from the best non-reference frame that does not require the image information of the preceding frame, thereby deteriorating image quality despite random editing. Provides the effect of preventing.

According to an arbitrary editing compressed video content providing system and a method according to an embodiment of the present invention, the unexpectedness of the edited video content by detecting a scene change point in the video content and editing the non-reference frame detected by the scene change to an arbitrary point It provides the effect of maintaining the context of the original video content while increasing the value.

According to another embodiment of the present invention and a method for providing a random edited compressed video content, the random point is randomly edited by randomly editing a GOP including an IDR frame as the best non-reference frame that does not affect the image quality of a preceding frame. In spite of the gap of the edited GOP, it is possible to prevent the deterioration of the quality of the following frame in the preceding frame.

According to another embodiment of the present invention and a method for providing a random edited compressed video content, a video section that is too short in the edited video content is defined by randomly arranging consecutive GOPs instead of a single GOP as an edited GOP array. To prevent them from being randomly distributed and to maintain the context of the original video content.

According to another exemplary embodiment of the present invention, a system and a method for providing a random edited compressed video content may include defining GOPs including the same VOP consecutively as an edited GOP array, thereby automatically editing the original content in units of a photographed object. It provides the effect of maintaining context.

Finally, according to another system and method for providing a random edited compressed video content according to another embodiment of the present invention, by defining the edited GOP array of GOPs containing the same person in succession, the original content through automatic random editing in the character unit Provides the effect of maintaining the context of

1 is a block diagram showing a system for providing arbitrarily compressed video content according to an embodiment of the present invention.

2 is a frame configuration diagram illustrating a GOP structure of compressed video content.

3 is a frame configuration diagram illustrating a GOP structure of arbitrarily edited compressed video content according to an embodiment of the present invention.

4 is a frame diagram illustrating a GOP structure of arbitrarily edited compressed video content according to another embodiment of the present invention.

5 is a GOP configuration diagram showing a GOP structure of compressed video content.

FIG. 6 is a GOP diagram illustrating a GOP structure of compressed video content arbitrarily edited in an edited GOP array unit according to another embodiment of the present invention. FIG.

7 is a GOP diagram illustrating a GOP structure of compressed video content arbitrarily edited in an edited GOP array unit according to another embodiment of the present invention.

8 is a flowchart illustrating a method for providing arbitrary edited compressed video content according to the present invention.

[Description of the code]

Video content providing system: 100 Video content storage: 110

Video content provider: 120 Video object analyzer: 130

Facial recognition unit: 140 User terminal: 200-1, 200-2

The terms or words used in this specification and claims are not to be construed as limiting in their usual or dictionary meanings, and the inventors shall refer to the principles on which terms and concepts may be defined in order to best explain their inventions. It should be interpreted as meanings and concepts corresponding to the technical spirit of the invention.

Therefore, the configuration shown in the embodiments and drawings described in this specification corresponds to a preferred embodiment of the present invention, and does not represent all of the technical spirit of the present invention, the configuration will be replaced at the time of filing of the present invention. There may be various equivalents and variations.

Throughout the specification, the description that a part “includes” an element means that the element may further include other elements, except for the absence of a special objection thereto. In addition, the terms “.. module”, “.. unit” and “.. system” described in the specification mean a unit that processes at least one function or operation, which is hardware or software or a combination of hardware and software. It may be implemented, and may be included in one device or each other device.

Hereinafter, a system for providing arbitrarily compressed video content (content_compress) according to the present invention will be described with reference to the accompanying drawings. 1 illustrates a system for providing arbitrary edited compressed video content (content_compress) according to an exemplary embodiment of the present invention. The random edited compressed video content (content_compress) includes a video content storage unit 110 and a video content providing unit 120.

According to an exemplary embodiment, the video content storage unit 110 and the video content providing unit 120 may be implemented in a single server system, or may be implemented as a server system configured as a separate server through a communication network. Further, according to the exemplary embodiment, hardware, such as a logic circuit, a memory, a storage device, or the like, may be implemented in the form of program codes of software that controls hardware, such as a logic circuit, a memory, a storage device, and the like, rather than a server-client system.

The video content storage unit 110 includes a first reference frame (frame_inter_first) that does not refer to the information of the preceding frame and a plurality of reference frames (frame_inter) which refer to the information of the preceding or following frames, respectively. A compressed video content (content_compress) including a GOP (GOP_1st) to an Nth GOP (GOP_Nth) (N is an integer of 2 or more) is stored.

2 illustrates a GOP structure of compressed video content (content_compress). A GOP is a set of frames and is composed of at least one reference frame (frame_inter) and at least one reference frame (frame_inter). The number of frames included in the GOP can be specified by the user. For example, the number of frames can be determined to have a time of about 0.5 seconds.

Since the reference frame (frame_inter) itself contains data such as brightness, color, etc. of all the pixels for the frame, the size of the data is large. According to a video compression standard such as MPEG, an I frame (Infra Frame) and an IDR frame (Instantaneous Decoder Refresh Frame) to be described later correspond to a reference frame (frame_inter).

Since the reference frame frame_inter includes only information about pixels or video object planes (VOPs) changed in the preceding reference frame (frame_inter), the data is small in size. Since the next frame in the video content is very short in time, the change of data between the successive frames is very small. Therefore, storing only the changed data is efficient in terms of data size. According to a video compression standard such as MPEG, P frames (Predicted Frame) and B frames (Bidirectional Frame) correspond to the reference frame (frame_inter). Since the P frame refers to the preceding I frame or another P frame, and the B frame refers to both the preceding frames and the following frames, it is common that the size of the data is smaller than the P frame.

In the case of the embodiment of FIG. 2, the k th GOP (GOP_kth) is composed of an I frame, a B frame, a B frame, a P frame, a B frame, and a B frame in order, an I frame that is a best-order reference frame (frame_inter_first), and a reference frame. It consists of four B frames and one P frame (frame_inter). In this case, the I frame is implemented through the data of the I frame when the video content is played. Then, the P frame, which is the third trailing frame in the I frame, refers to the I frame, and then the B frame, which is the second trailing frame in the I frame. The video content is played back in such a way that it refers to the preceding I frame and the following P frame, and then the B frame, which is the third trailing frame in the I frame, refers to the preceding I frame and B frame and the following P frame.

Meanwhile, the video content storage unit 110 of the present invention may be implemented to include a function of converting and storing uncompressed video content into compressed video content (content_compress), or to store compressed video content (content_compress) that is already compressed. May be

The video content providing unit 120 provides the compressed video content (content_compress) to the user terminals 200-1 and 200-2 in units of GOPs through a communication network, and at least some of the GOPs are in order. 1, 200-2), and precedes k-th GOP (GOP_kth) (k is an integer of 2 or more and N-1 or less) k-1 GOP (GOP_k-1th) or k + 1 GOP (GOP_k). + 1th) to the user terminal (200-1, 200-2) in any order that follows.

The video content providing unit 120 provides a video content through a communication network such as the Internet or an intranet in the case of a server-client system, and provides a video content through system internal data processing rather than a communication network when implemented on a single client. can do.

The video content providing unit 120 provides at least some of the GOPs to the user terminals 200-1 and 200-2 in the order of the original video content, and at least some other GOPs are different from the order of the original video content. To (200-1, 200-2). In this case, the ratio of the non-edited GOPs provided in the order and the edited GOPs provided in the order may be variously applied according to an embodiment. For example, if you want to make a lot of changes in the original video content, the ratio of edited GOP to non-edited GOP may be 90% .In contrast, if you want to make some changes while maintaining the context of the original video content, The ratio can be set at 10%. According to an exemplary embodiment, the compressed video content (content_compress) providing system may determine the ratio of the edited GOP to the non-edited GOP at an arbitrary ratio.

In order to eliminate the need to specify additional editing GOPs by the user, the determination of the GOPs to be edited GOPs is automatically performed by the system for providing compressed video content (content_compress) rather than user decision or by a predetermined algorithm through image processing. Configure to decide.

Any kth GOP (GOP_kth) may be disposed adjacent to or in front of a preceding k-1 GOP (GOP_k-1th) as shown in FIG. 3, for example, a k-20 GOP (GOP_k-20th). It may be spaced apart from the k-1th GOP (GOP_k-1th) in front of the. Conversely, any k GOP (GOP_kth) is arranged adjacent to or behind the following k + 1 GOPs (GOP_k + 1th) as shown in FIG. 4, for example k + 20 GOP (GOP_k + 20th). ) May be spaced apart from the k + 1th GOP (GOP_k + 1th) at the rear of the.

Meanwhile, the video content providing unit 120 may randomly edit a single compressed video content (content_compress) or may arbitrarily edit a plurality of compressed video contents (content_compress). When randomly editing a plurality of compressed video contents (content_compress), the first compressed video content (content_compress_1st) is composed of 150 GOPs, and the second compressed video content (content_compress_2nd) is composed of 300 GOPs. If the third compressed video content (content_compress_3rd) is composed of 200 GOPs, the video content providing unit 120 regards the compressed video content (content_compress) composed of a total of 650 GOPs according to an arbitrary compressed video content order and edits randomly. This may be provided to the user terminals 200-1 and 200-2. According to this embodiment, the unexpectedness of the randomly edited video content is increased.

On the other hand, the recent video compression technology has been developed to position the scene reference point (frame_inter_first) of the GOP through the scene change detection through image processing. This is because when the scene transformation occurs in the middle of the GOP, frame correlation between the preceding frame and the following frame is lowered, and thus, there is a problem of increasing the data amount of the B frame or P frame, which is the reference frame (frame_inter). . In the embodiment of the present invention, by focusing on this point, the scene change point is edited by scene unit by using the GOP that defines the transition point as the best reference frame (frame_inter_first) as the editing GOP through scene change detection, so that it is edited in units of scenes. It can provide the effect of automatically generating content. To this end, the video content providing unit 120 may be configured to select a GOP including the I frame located at the scene change point detected through scene change detection as the best reference frame (frame_inter_first) as the k th GOP (GOP_kth). Do.

Meanwhile, a method of arbitrarily editing video content and providing the same to a user may be provided. First, the video content providing unit 120 may change the order of the GOP without changing the compressed video content (content_compress) file and change the order of the GOP to provide a video transmission format. Second, the video content providing unit 120 changes the GOP order of the compressed video content (content_compress) file and stores the compressed video content (content_compress) file according to an arbitrary editing result or compresses the video according to the changed GOP order in the buffer memory. A copy of all or part of the content (content_compress) file may be uploaded and provided.

In the former case, there is no problem. In the latter case, the following data loss problem may occur according to the change of the GOP order. As shown in FIG. 2, the last frame of the k-1 GOP (GOP_k-1th), which is the frame immediately before the I frame that is the best reference frame (frame_inter_first) of the k GOP (GOP_kth), is the bidirectional reference frame (frame_inter). B frame. As shown in FIG. 3, if the k-th frame GOP_kth is moved to another position as a result of the arbitrary editing of the video content providing unit 120, the B frame, which is the last frame of the k-1 GOP GP_k-1th, is referred to. The trailing frame may be lost and the frame image may not be completely created.

According to an embodiment of the present invention, in order to solve the above problem, data in some frames is selected by selecting a GOP including the IDR frame, which is not referenced by the preceding frames of the preceding GOP, as the best-order non-reference frame as the kth GOP (GOP_kth). It can provide the effect of preventing the loss.

On the other hand, a single GOP is very short, such as a few tenths or a few seconds, so if only a single GOP is randomly edited and moved to another location, it may not be possible to convey the context of the original video content and may be perceived as noise to the user. There is a problem. In order to prevent such a problem, the video content providing unit 120 according to an embodiment of the present invention is edited as defined by k th GOP (GOP_kth) to k + m GOP (GOP_k + mth) (m is an integer of 10 or more). The user terminals 200-1 and 200-2 in any order to precede the GOP array array_GOP_edit or to follow the k-1 m GOP (GOP_k-1 th) or the k + m + 1 GOP (GOP_k + m + 1 th). It is desirable to provide to. In this case, at least 11 GOPs should be set to one edit GOP array (array_GOP_edit). Since the object to be arbitrarily edited has a length of at least 1 second to 10 seconds, m is preferably set to 10 or more.

5 is a GOP diagram illustrating a GOP structure of compressed video content (content_compress), and FIGS. 6 and 7 are compressed video content (content_compress) arbitrarily edited in an edited GOP array (array_GOP_edit) according to an embodiment of the present invention. A GOP configuration diagram showing the GOP structure of the. According to this embodiment, the video content providing unit 120 defines the third GOP (GOP_3rd) to the fifth GOP (GOP_5th) as an edited GOP array (array_GOP_edit) (k = 3, m = 2). According to an embodiment, as shown in FIG. 6, the editing GOP array array_GOP_edit is arranged to be adjacent to a second GOP (GOP_2nd) which is an adjacent preceding GOP, or as shown in FIG. 7, the editing GOP array (array_GOP_edit) may be The sixth GOP (GOP_6th), which is an adjacent trailing GOP, may be disposed to be trailed apart from each other.

According to this embodiment, since the edited GOP array (array_GOP_edit) of at least 10 consecutive GOPs are edited to provide users with randomly edited video content that can increase the unexpectedness while maintaining the context of the original video content. It is effective.

On the other hand, the editing GOP arrays (array_GOP_edit) is more preferably composed of GOPs containing the same video object plane (VOP), such as the same person, the same object, the same background. To this end, the random edited compressed video content (content_compress) providing system includes a video object analyzer 130 for analyzing VOPs (Video Object Planes) of the image of the best reference frame (frame_inter_first) and the reference frame (frame_inter). In this case, the video content providing unit 120 may be configured to define a kth GOP (GOP_kth) to a k + m GOP (GOP_k + mth) including the same VOP consecutively as an edited GOP array (array_GOP_edit).

In another embodiment, it is preferable that the editing GOP arrays (array_GOP_edit) are composed of GOPs that continuously include the same person through face recognition. To this end, the arbitrary editing compressed video content (content_compress) providing system recognizes the face of the person included in the image of the best line reference frame (frame_inter_first) and the reference frame (frame_inter) and associates it with the face identifier (id_facial). More). In this case, the video content providing unit 120 defines k-th GOP (GOP_kth) to k + m (GOP_k + mth) GOPs continuously associated with the same face identifier (id_facial) as an edited GOP array (array_GOP_edit). .

At this time, the facial identifier (id_facial) is not particularly limited, and the name, alias, and identifier of a previously recognized face and its corresponding person can be found through the internal memory or an external server, and used as the face identifier (id_facial). It may be an identifier assigned in a specific order or arbitrarily according to facial features.

Hereinafter, a method for providing arbitrary edited compressed video content (content_compress) according to the present invention will be described with reference to FIG. 8.

First, the video content providing system 100 includes a best reference frame (frame_inter_first) that does not refer to the information of the preceding frame and a plurality of reference frames (frame_inter) which refer to the information of the preceding or following frames, respectively. A first step S10 of storing compressed video content (content_compress) including first GOP (GOP_1st) to Nth GOP (GOP_Nth) (N is an integer of 2 or more) is performed.

In this case, the reference frame (frame_inter_first) may be an I frame or an IDR frame, and the reference frame (frame_inter) may be a B frame or a P frame.

Next, a second step (s20) in which the video content providing system 100 receives a random edit request for requesting arbitrary editing of the compressed video content (content_compress) from the user terminals 200-1 and 200-2 through a communication network. Perform

Finally, the video content providing system 100 provides the compressed video content (content_compress) to the user terminals 200-1 and 200-2 in units of GOPs through a communication network, and at least some of the GOPs are provided in order. 200-1, 200-2, and precede kth G-1 GOP (GOP_k-1th) with kth GOP (GOP_kth), where k is an integer of 2 or more and N-1 or less. A third step s30 is provided to the user terminals 200-1 and 200-2 in any order following GOP_k + 1th.

In the third step s30, the video content providing system 100 converts the I frame located at the scene change point detected by the scene change detection into the best reference frame (frame_inter_first). It is preferable to select the containing GOP as the kth GOP (GOP_kth).

In order to prevent data loss due to the movement of the best-order reference frame (frame_inter_first) of the k-th GOP (GOP_kth) prior to the k-th GOP (GOP_kth), In step S30, the video content providing system 100 selects a GOP including an IDR frame (Instantaneous Decoder Refresh) as the best-order reference frame (frame_inter_first) as the k th GOP (GOP_kth).

In the third step s30, the video content providing system 100 performs k GOP (GOP_kth) to k + m GOP (GOP_k + mth) (m is an integer of 10 or more) so that the editing GOPs which are automatically edited have an appropriate length. Is defined as the editing GOP array (array_GOP_edit), and the arbitrary order to precede the editing GOP array (array_GOP_edit) to k-1 GOP (GOP_k-1th) or to k + m + 1 GOP (GOP_k + m + 1th). It is desirable to provide to the user terminal (200-1, 200-2).

In an embodiment of auto editing in an edited GOP array (array_GOP_edit) unit, the first step (s10) is performed by the video content providing system 100 to edit GOPs including the same VOP consecutively in an edited GOP array (array_GOP_edit) unit. The method further includes a first substep s11 for analyzing the VOPs (Video Object Planes) of the image of the best row reference frame (frame_inter_first) and the reference frame (frame_inter), and the second step (s20) includes a VOP having the same video content system. It is preferable to define the kth GOP (GOP_kth) to k + m GOP (GOP_k + mth) including the consecutively as an edited GOP array (array_GOP_edit).

In an embodiment of auto editing in an edited GOP array (array_GOP_edit) unit, the first step (s10) is performed by the video content providing system 100 to edit GOPs containing the same person consecutively in an edited GOP array (array_GOP_edit) unit. And further including a second substep s12 for recognizing the face of the person included in the image of the best row reference frame frame_inter_first and the reference frame frame_inter and associating it with the face identifier id_facial. Preferably, the video content system defines a kth GOP (GOP_kth) to a k + m GOP (GOP_k + mth) continuously associated with the same face identifier (id_facial) as an edited GOP array (array_GOP_edit).

The description of the invention herein describes the preferred embodiment, and the invention is not limited to this embodiment. Those skilled in the art to which the present invention pertains can make various changes and modifications to the above embodiments without departing from the technical spirit of the present invention, the technical idea of the present invention is to make all such various changes and modifications Include.

Claims

The first GOP to the Nth GOPs each configured to include a best-order non-reference frame not referring to the information of the preceding frame and a plurality of reference frames referencing the information of the preceding or following frame (where N is an integer of 2 or more). A video content storage unit for storing the compressed video content; And

The compressed video content is provided to the user terminal in units of GOPs through a communication network, and at least some of the GOPs are provided to the user terminal in order, and kth GOP (k is an integer of 2 or more and N-1 or less) is made. and a video content providing unit provided to the user terminal in any order preceding the k-1 GOP or following the k + 1 GOP.
The method of claim 1,

And the best-order dereference frame is an I frame, and the reference frame is a P frame and / or a B-directional frame.
The method of claim 2, wherein the video content providing unit,

And a GOP including the I frame located at the scene change point detected through scene change detection as the k-th non-reference frame, as the k-th GOP.
The method of claim 2, wherein the video content providing unit,

And a GOP including an InstantRane Decoder Refresh (IDR) frame as the best non-reference frame as the k-th GOP.
According to claim 1, The video content providing unit,

The user terminal in any order to precede the k-1 GOP or to follow the k + m + 1 GOP to the editing GOP array defined by the k th GOP to the k + m GOP (m is an integer of 10 or more) Random edited compressed video content providing system, characterized in that provided.
The system of claim 5, wherein the random edited compressed video content providing system comprises:

And a video object analyzer configured to analyze video object planes (VOPs) of the best-order non-reference frame and the image of the reference frame.

And the video content providing unit defines the k-th GOP to the k + m GOP including consecutively the same VOP as the edited GOP array.
The system of claim 5, wherein the random edited compressed video content providing system comprises:

And a face recognizing unit recognizing a face of a person included in the image of the reference frame and the best non-reference frame and associating it with a face identifier.

And the moving picture contents providing unit defines the k th GOP to the k + m GOPs associated with the same face identifier consecutively as the edited GOP array.
The video content providing system includes a first GOP to an Nth GOP, each of which includes a plurality of reference frames that refer to information of a best-order non-reference frame that does not refer to information of a preceding frame and information of a preceding or following frame ( N is an integer of 2 or more); a first step of storing the compressed video content;

A second step of the video content providing system receiving a random edit request for requesting arbitrary edit of the compressed video content from a user terminal through a communication network; And

The video content providing system provides the compressed video content to the user terminal in units of GOPs through a communication network, wherein at least some of the GOPs are provided to the user terminal in order, and the kth GOP (k is equal to or greater than 2). An integer less than or equal to N-1) a third step of providing the user terminal in any order preceding the k-1 GOP or following the k + 1 GOP; How to Provide.
The method of claim 8, wherein the first step,

The video content providing system stores the compressed video content, wherein the best-order non-reference frame is an I frame, and the reference frame is a P frame and / or a bidirectional frame. Random edited compressed video content providing method characterized in that.
The method of claim 9, wherein the third step,

The video content providing system provides a randomized edited video content including a GOP including an I frame located at a scene change point detected through scene change detection as the best non-reference frame as the k th GOP. Way.
The method of claim 9, wherein the third step,

And the video content providing system selects a GOP including an Instantaneous Decoder Refresh (IDR) frame as the best non-reference frame as the k-th GOP.
The method of claim 8, wherein the third step,

The video content providing system defines the k th GOP to the k + m GOP (m is an integer of 10 or more) as an edited GOP array, and the edited GOP array precedes the k-1 GOP or k + m. And providing the user terminal in any order so as to follow the +1 GOP.
The method of claim 12,

The first step may further include a first sub-step of the video content providing system analyzing VOPs (Video Object Planes) of the best non-reference frame and the image of the reference frame.

In the second step, the video content system defines the kth GOP to the k + m GOP including the same GOP consecutively as the edited GOP array, characterized in that the edited GOP array method.
The method of claim 12,

The first step may further include a second sub-step of the video content providing system recognizing a face of a person included in an image of the best non-reference frame and the reference frame and associating it with a face identifier.

In the second step, the video content system defines the kth GOP to the k + m GOPs associated with the same face identifier consecutively as the edited GOP array.