KR20200022640A - System for selecting segmentation video using high definition camera and the method thereof - Google Patents

Abstract

The present invention relates to a segmented video selection system using a high-definition camera which segments a single full video for a busking performance photographed with a high-definition camera into a plurality of segmented videos and selects a video to be played based on big data probabilities, and a method thereof. The segmented video selection system using the high-definition camera can select an optimal video to be played at every moment by considering a music flow, a busking lighting, a busking space, and a busking background among the plurality of segmented videos to transmit the same as a highlight video. The segmented video selection system using the high-definition camera comprises a video receiving part, a screen segmentation part, and a selection part.

Description

Segmentation image selection system using high quality camera and its method {SYSTEM FOR SELECTING SEGMENTATION VIDEO USING HIGH DEFINITION CAMERA AND THE METHOD THEREOF}

The present invention relates to a split image selection system and a method using a high definition camera, and more particularly, a single full image of a busking performance shot by a high definition camera is divided into a plurality of split images to generate a big data. A system for selecting a play video based on probability.

Recently, with the development of busker culture, buskers perform various genres of music on the street through various repertoires. In addition, due to the spread of smart devices, an environment in which a user of the smart device can acquire various information on the go has been provided.

Currently, there is no system for acquiring information on a busker using a smart device. Therefore, a system capable of providing information on the busker is required for the promotion of the performance from the perspective of the busker and to obtain the performance information from the perspective of the audience.

Korean Patent No. 10-1271590, which is a prior invention, relates to a smart real-time performance system and a method thereof. The present invention receives a demo image taken by a plurality of performance participation candidates and stores in the image DB, the image management unit for providing the demo image or the performance image to a user terminal of a third party, and the demo input from the third party A preference evaluation unit for evaluating preferences for the plurality of performance participation candidates using preference information on a video, a performance participant determination unit determining a performance participation candidate having the highest preference evaluation information as a performance participant, and the third Disclosed is a smart real-time performance system including a performance score calculator for calculating a performance score for the performance participant using the performance evaluation information of the performance participant inputted from the user.

However, the preceding invention provides a demo image to a service user and suggests a method of constructing a performance using preference information, but there is a limit to applying the prior invention to buskers performing various performances in the street. have.

Furthermore, in order to produce existing performance broadcasting images, each group busker requires a camera for each person, switcher equipment, and audio / video equipment. Therefore, there is a limit of increase in cost and manpower according to essential equipment.

An object of the present invention is to provide a system that can provide a full screen or close-up screen for a busking performance by producing a plurality of camera effects using a single high-definition camera.

In addition, an object of the present invention is to provide a technology that can solve the cost and quality issues of the production of the performance broadcast by selecting and transmitting the optimal image suitable for the busking environment of the plurality of divided images.

In addition, an object of the present invention is to provide a technology that can transmit the highlight image by selecting the optimal play image at every moment in consideration of the music flow, busking lighting, busking space and the background of the plurality of divided images. do.

In the split image selection system using a high quality camera according to an embodiment of the present invention, an image receiving unit for receiving a single full image of a busing performance photographed using a high quality camera, and object recognition of a screen of the full image And a screen divider for dividing the full image into a plurality of divided images through a process, and a selector to select a play image based on a probability of big data according to a busking environment among the plurality of divided images.

The image receiving unit may receive the single full image of the busking image captured by the single high quality camera.

The screen splitter may extract an object of a person participating in the busking performance and an object of the musical instrument used for the busking performance, and divide the screen splitter into the plurality of divided images according to the number of extracted objects.

The screen splitter may divide the extracted person or musical instrument into the plurality of divided images including the overlapped objects.

The selector may select the play image among the plurality of divided images sampled using the big data learned for each music.

The selection unit extracts at least one sampling of the busking environment from at least one of a music flow, a busking illumination, a busking space, and a busking background from the divided image, and compares an extraction result with the big data to divide the plurality of divisions. The image may be classified into a normal image or an issue image.

The selector may calculate a probability value for each of the plurality of divided images by applying a machine learning technique to the extracted sampling for the busking environment.

The selector may select the play video by combining at least one sound feature of multitrack sensing, sound analysis, and BPM recognition with the probability value.

The divisional image selection system using a high-definition camera according to an embodiment of the present invention further includes a control unit for controlling the selection of the play image and the selection of the play image based on probability among the plurality of divided images based on a participant's opinion. can do.

In a method of operating a system for selecting a divided image for a busking performance using a high quality camera according to an embodiment of the present invention, a single full image of the busking performance photographed using the high quality camera Receiving the step, the step of dividing the full image into a plurality of divided images through the object recognition process for the screen of the full image and the play image based on the probability of big data according to the busking environment of the plurality of divided images Selecting.

The dividing into the plurality of divided images may include extracting objects of a person participating in the busking performance and an instrument used for the busking performance, and splitting the plurality of divided images into the plurality of divided images according to the number of extracted objects.

The selecting of the play image may select the play image from among the plurality of divided images sampled using the big data learned for each music.

The selecting of the play image may include extracting a sampling of at least one of the bursting environment of music flow, bursting illumination, bursting space, and bursting background from the split image, and comparing the extraction result with the big data. Classifying the plurality of divided images into a general image or an issue image, and calculating a probability value for each of the plurality of divided images by applying a machine learning technique to the extracted sampling for the busking environment And selecting the play video by combining at least one sound feature of multitrack sensing, sound analysis, and BPM recognition with the probability value.

According to an embodiment of the present invention, by using a single high-definition camera to produce a plurality of camera effects, it is possible to provide a full screen or a close-up screen for the busking performance.

In addition, according to an embodiment of the present invention, by selecting an optimal image suitable for a busking environment from among a plurality of divided images and transmitting it, it is possible to solve the cost and quality issues of the production of a performance broadcast.

In addition, according to an embodiment of the present invention, in consideration of music flow, bursting lighting, bursting space, and bursting background among a plurality of divided images, an optimal play image may be selected at every moment and a highlight image may be transmitted.

1 is a block diagram illustrating a configuration of a split image selection system using a high quality camera according to an exemplary embodiment of the present invention.
2 is a block diagram of a split image selection system using a high-definition camera according to an embodiment of the present invention.
3 illustrates a conceptual example of a split image according to an exemplary embodiment of the present invention.
4 illustrates an example of an object recognition process according to an embodiment of the present invention.
5 illustrates an example of classifying an issue image and a general image according to an embodiment of the present invention.
6 illustrates an example of selecting a play video according to an embodiment of the present invention.
7 illustrates an example of a table for selecting a play video according to an embodiment of the present invention.
8 illustrates an example of transmission of a play video according to an embodiment of the present invention.
9 is a flowchart illustrating a method of selecting a divided image using a high quality camera according to an embodiment of the present invention.

Hereinafter, exemplary embodiments of the present invention will be described in detail with reference to the accompanying drawings. However, the present invention is not limited or limited by the embodiments. Also, like reference numerals in the drawings denote like elements.

In addition, terms used in the present specification (terminology) are terms used to properly express preferred embodiments of the present invention, which may vary depending on the intention of the viewer, the operator, or customs in the field to which the present invention belongs. Therefore, the definitions of the terms should be made based on the contents throughout the specification.

The present invention shoots a busking performance with a single camera, divides several scenes of a single recorded full video, and transmits the most important highlight scene to a live broadcast. We propose a system for solving cost and quality issues.

Embodiments of the present invention are divided into a plurality of divided images through the object recognition process using a single full image of the busking performance photographed using a single high-definition camera, based on the probability of big data according to the busking environment The present invention relates to a technology for selecting and transmitting a play video by using a single high-definition camera, and by providing a plurality of camera effects through a single high-definition camera, providing a high-quality video at a low budget to perform producers, buskers (or performers, performers, singers, etc.). ) And participants (a viewer or a video viewer watching a busking performance) to provide a technology that can generate revenue.

The split image selection system and method using the high quality camera according to the embodiments of the present invention will be described in detail with reference to FIGS. 1 to 9.

1 is a block diagram illustrating a configuration of a split image selection system using a high quality camera according to an exemplary embodiment of the present invention.

Referring to FIG. 1, a split image selection system 100 using a high definition camera according to an embodiment of the present invention divides a single full image of a busking performance taken by a high definition camera into a plurality of split images. The play video is selected based on the probability of the big data.

Accordingly, the split image selection system 100 using the high quality camera according to the exemplary embodiment of the present invention includes an image receiver 110, a screen divider 120, and a selector 130.

The image receiver 110 receives a single full image of a busing performance taken using a high quality camera.

For example, when a busker (street performer) performs a street performance, the performance producer shoots a busking performance using a single high-definition camera, and the image receiving unit 110 receives a busking performance from the performance producer. A single full image can be received.

According to an embodiment, the performance producer is a person who shoots a full image of a busking performance, which may be a staff, a busker, or a viewer, and a host application installed in a mobile device having a full image taken from a high-definition camera. Can be sent via In this case, the high definition camera may be a UHD (Ultra-HD) 4320P / 4K camera, and the image receiving unit 110 may receive a single 4K full image captured by the high definition camera.

The screen divider 120 divides the full image into a plurality of divided images through an object recognition process on the screen of the full image.

The screen divider 120 may divide a single 4320P / 4K full image received from the image receiver 110 into four 1080P images, but the number of divided images is not limited to four.

For example, the screen splitter 120 may use a face (face) or object recognition algorithm to select a person participating in a busking performance on a screen (instantaneous frame) and an object of a musical instrument used for the busking performance. The extracted image may be divided into a plurality of divided images according to the number of objects of the extracted person or musical instrument.

According to an embodiment, when there is a person who plays only a musical instrument, a singer, or a person who sings a musical instrument in a busking performance, the screen divider 120 first extracts a person and each object of the musical instrument. The person who plays the song, the person who sings, and the person who sings with the instrument can be classified and divided into divided images for each.

According to another embodiment, the screen splitter 120 of the split image selection system 100 using the high-definition camera according to an embodiment of the present invention, in the process of recognizing the object in the full image, the overlapping person and the instrument or microphone It can also be extracted. For example, if participants A, B, C and instruments D, E, and microphone F are extracted, if participants A, instruments E, microphone F overlap, and participants B, C overlap, the present invention provides participants A, The instrument E and the microphone F may be divided into one image, the participants B and C may be divided into one image, and the instrument D may be divided into one image to divide a plurality of divided images. That is, when overlapping of objects occurs, the screen splitter 120 divides the image by determining the overlapping objects as one group.

However, in the busking performance, the viewer may sing or clap in a song, so the screen divider 120 according to the embodiment of the present invention preferentially selects an object which is clearly a busker such as an instrument and a microphone. Can be extracted. In addition, the split image selection system 100 using a high-definition camera according to an embodiment of the present invention may extract an object of a person or a musical instrument corresponding to the number of buskers to be input.

In this case, since a recognition algorithm used in the screen divider 120 uses an algorithm used to recognize a face, an object, or a voice in the existing technology, a detailed description thereof will be omitted. In addition, the screen dividing unit 120 may apply a machine learning technique or an artificial intelligence (AI) technique in addition to an algorithm for extracting a person or a musical instrument from a screen of a full image, and is not limited to the algorithm.

The selector 130 selects a play video based on a probability of big data according to a busking environment among a plurality of divided images.

For example, the divided image selection system 100 using a high-definition camera according to an embodiment of the present invention may store and maintain big data learned for each music, and download or download big data from an external storage device or an external server. You can update it. In this case, the big data is data obtained by learning sound sources for all music pieces, and may be classified into songs or musical instruments and dataized.

Accordingly, the selector 130 may select a play image from data sampled for each of the plurality of divided images divided by the screen divider 120 based on the big data.

For example, the selector 130 compares sampling and big data for each of the plurality of divided images to select a climax or highlight play image according to a music flow, or to play a guitar, a drum or a violin, or the like. If determined to be, the play video representing the person playing the instrument can be selected. In addition, the selector 130 may select a play image representing a person or an object of a musical instrument to which a busing light such as a pin light is pointing, or select a play image representing a busing space such as indoors or outdoors. A play video representing the background of the busking, such as, dry pole or Han river, can be selected.

The selection of the selection unit 130 may be performed based on big data including all music pieces, or may be based on a preset selection input of a manager (or user) or a selection of participants.

As another example, the selector 130 extracts a sampling of at least one or more busking environments of a music flow, a busking lighting, a busking space, and a busking background from a divided image, and compares the extraction result with big data. A plurality of divided images may be classified as general images or issue images. Thereafter, the selector 130 applies a machine learning technique to the sampling extracted for the busking environment to calculate probability values for each of the plurality of divided images, multitrack sensing, sound analysis, A play video may be selected by combining at least one sound feature among BPM recognitions.

According to an embodiment, the split image selection system 100 using a high-definition camera according to an embodiment of the present invention is a music flow and musical instruments that are preferred by a plurality of participants in all music pieces, highlighting busking lights, busking spaces, and buses. The king background information may be stored and maintained in the big data, and the selector 130 may classify the general image and the issue image by comparing sampling and big data for each of the plurality of divided images based on the information. . Thereafter, the selector 130 may calculate a probability value for each issue image by applying a machine learning method based on the issue image, and multitrack sensing, sound analysis, and BPM recognition based on a musical instrument and a music flow based on the probability value. At least one of the sound features may be combined to select an image representing the highest probability value as the play image.

That is, the selector 130 may be configured according to the climax or the highlight of the busking performance among the plurality of divided images based on at least one of the busking environment of the music flow, the busking lighting, the busking space, and the busking background. It is characterized by selecting a play video.

In this case, according to an embodiment, the busking environment is a repertoire of a busking performance, a performance venue, a latitude, longitude, an address, a noise level, an electrical support, an indoor performance or an outdoor performance, time information, and a performance of a performance venue. It may further include information of at least one of the order.

In addition, as shown in Figure 1, the split image selection system 100 using a high-definition camera according to an embodiment of the present invention, the selection of the play image based on the opinions of the participants, and the probability-based play of the plurality of divided images The control unit 140 may further control the selection of an image.

For example, the controller 140 may drive a play image selected every time in seconds and minutes based on a climax or highlight for each of the plurality of divided images, and from a mobile device possessed by a participant. The play video may be selected and expressed based on the selection input of the participants.

In addition, the controller 140 may control the stream process based on the image quality, size, and type of the full image received from the image receiver 110, and may be based on big data that is input or updated by a participant. The screen divider 120 may be controlled. That is, the controller 140 may control overall operations such as driving, performing, and expressing the components of the divided image selection system 100 using the high quality camera according to the exemplary embodiment of the present invention.

In addition, the controller 140 may request information of the busker from the producer or the busker and use the information according to the response to divide the screen or select the play video. According to an embodiment, the information of the busker may include a team name of the busker, a list of team members, a gender of a team member, a performance category, a genre, album information, a recommender, a position for each team member, a total number of team members, and a uniform resource locater. URL), a URL of Facebook, a URL of Twitter, a URL of a fan community, an active country, an active city, and a tag.

2 is a block diagram of a split image selection system using a high-definition camera according to an embodiment of the present invention.

2, the split image selection system 200 using the high definition camera according to the embodiment of the present invention divides the full image received from the host application 220 (b) using the image production process 230. , Select play video from split video.

More specifically, the performance producer 210 basically has a single high-definition camera 4K camera and a mobile device, using a single high-definition camera to shoot (a) the image of the busking performance, and within the mobile device The captured full image may be uploaded to the installed host application (eg, BuskingPlay 220).

Accordingly, the divided image selection system 200 using the high-definition camera according to the embodiment of the present invention receives the full image uploaded to the host application 220 (b) and uses the image production process 230 to obtain the full image. The screen may be divided into a plurality of divided images, and a play image may be selected from the plurality of divided images and transmitted to the participant application 250 (d). Here, the image production process 230 may include a screen divider 120 and a selector 130 according to the embodiment of the present invention shown in FIG. 1.

At this time, the segmentation image selection system 200 using a high-definition camera according to an embodiment of the present invention drives and represents an algorithm 240 such as a device protocol, a machine learning technique, and a streaming server technique in the image production process 230. c), the full video is divided into a plurality of divided images, and a play video can be selected from the divided images.

Here, the host application 220 and the participant application 250 is an application installed on the mobile device possessed by the performer 210 and the participant, the mobile device is a personal computer (PC), a laptop computer (laptop computer), It is a terminal device possessed by producers, buskers, and participants such as smartphones, tablets, and wearable computers. Service screen composition and data under the control of a web / mobile site or a dedicated application. The overall service operation can be performed such as input, data transmission, and data storage. In addition, the mobile device may refer to an application downloaded and installed in the mobile device.

In addition, the image production process 230 is a software application software (SW), a scene area automatic setting function for classifying a full image into a plurality of divided images, and a climax based on sampling of each of the plurality of divided images. Or a highlight scene selection function for extracting a highlight, and an on-air function for uploading and providing a selected play video to a chat room accessed by a plurality of participants through the participant application 250.

3 illustrates a conceptual example of a split image according to an exemplary embodiment of the present invention.

Referring to FIG. 3, a split image selection system using a high definition camera according to an embodiment of the present invention captures a busking performance using a single high definition camera 310, and divides a single full image into a plurality of split images. do.

In this case, the single high-definition camera 310 is a UHD (Ultra-HD) 4320P / 4K camera, and can capture a single 4K full image for a busking performance.

The split image selection system using a high definition camera according to an exemplary embodiment of the present invention may receive an input stream of a single 4K full image from a single high definition camera 310. In this case, the received full image of 4320P / 4K is a combination of four 1080P FULL HD screens, and as shown in FIG. 3, may be divided into four divided images. However, the number, shape, and size of the divided images are not limited thereto.

4 illustrates an example of an object recognition process according to an embodiment of the present invention.

Referring to FIG. 4, the split image selection system using a high definition camera according to an embodiment of the present invention applies an object recognition process for splitting a full image received from a high definition camera into a plurality of split images.

For example, a split image selection system using a high definition camera according to an embodiment of the present invention receives a full image as shown in FIG. 4 (a) from a high definition camera, and performs an object recognition process on a screen (instantaneous frame) among the full images. A person participating in the busking performance as shown in FIG. 4 (b) may be extracted (red square).

In this case, the object recognition process is a recognition algorithm, which recognizes a face (face), an object, or a voice by using big data and machine learning, and is a process used in the existing technology. Detailed description is omitted. However, the present invention is not limited to the above-described object recognition process, since a library such as OpenCV (Open Source Computer Vision) may be applied to recognize a face, an object, or a voice on a screen.

In addition, the split image selection system using a high-definition camera according to an embodiment of the present invention is a musical instrument, a microphone, etc. used for the busking performance in addition to the recognition of the person (face) participating in the busking performance as shown in FIG. You can also extract more objects from.

For example, the split image selection system using the high-definition camera according to an embodiment of the present invention may extract an overlapping person and a musical instrument or a microphone in the process of recognizing the object in the full image. According to an embodiment, if participants A, B, C and instruments D, E, and microphone F are extracted, participants A, musical instrument E, microphone F overlap, and participants B, C overlap, the present invention provides participants A, The instrument E and the microphone F may be divided into one image, the participants B and C may be divided into one image, and the instrument D may be divided into one image to divide a plurality of divided images. That is, the present invention is based on segmenting the image by judging the overlapping objects as one group when the overlapping of objects occurs.

5 illustrates an example of classifying an issue image and a general image according to an embodiment of the present invention.

In the split image selection system using a high quality camera according to an exemplary embodiment of the present invention, as shown in FIGS. 3 and 4, a split image is divided into a plurality of split images according to objects extracted through an object recognition process from a full image received from a high quality camera. The divided image may be classified into an issue scene 541 or a general scene 542.

Referring to FIG. 5, a split image selection system using a high-definition camera according to an exemplary embodiment of the present invention samples each of a plurality of split images 510 and divides the splits sampled using the big data 530 in the classifier 520. The image 510 may be classified into an issue scene 541 and a general scene 542.

The classifier 520 extracts at least one or more busking environments among music flows, busking lights, busking spaces, and busking backgrounds from the sampled divided image 510, and compares the extracted result with the big data 530 to divide the divided results. The image 510 may be classified into an issue image (or issue scene 541) and a general image (or general scene 542).

In this case, the big data 530 may be stored and maintained in a split image selection system using a high quality camera according to an embodiment of the present invention, and may be information downloaded or updated from an external storage device or an external server. . In addition, the big data 530 is data obtained by learning sound sources for all music, and may be classified into songs or musical instruments and dataized.

For example, the big data 530 may include information about music flows and climax or highlights of the instruments preferred by a plurality of participants for each of the entire music pieces, as well as highlighted busking lights, busking spaces, and busking backgrounds. Information can be stored and maintained, and can be updated in real time from an external storage device or an external server. Accordingly, the split image selection system using the high-definition camera according to an embodiment of the present invention extracts the highlight, climax or highlight portion for each of the divided images 510 sampled based on the big data 530. It may be classified as an issue image 541. In addition, the music flow (or performance) other than the emphasis, climax, or highlight portion in the music flow may be classified into the general image 542.

That is, the divided image selection system using a high-definition camera according to an embodiment of the present invention is a plurality of issue images 541 and a plurality of images according to a highlight, climax or highlight portion in a busking performance playing a piece of music. The general image 542 may be classified, and the classified issue image 541 and the general image 542 may be changed, corrected, and classified at every second or minute.

6 illustrates an example of selecting a play video according to an embodiment of the present invention.

Referring to FIG. 6, the split image selection system using the high-definition camera according to an embodiment of the present invention performs sampling 620 on the plurality of split images 610 and performs sampling by applying a machine learning technique. A probability value for the split image 620 may be calculated. Next, the present invention selects an image representing the highest probability value as the play image 640 by combining at least one or more sound features 630 of instrument and music flow-oriented multitrack sensing, sound analysis, and BPM recognition. Can be.

For example, the segmentation image selection system using a high-definition camera according to an embodiment of the present invention applies a machine learning method to the segmented image 620 sampled based on big data, such as music flow, busking lighting, busking space, and the like. A probability value according to at least one or more busking environments of the bursting background may be digitized. Then, the split image selection system using a high-definition camera according to an embodiment of the present invention is an instrument (eg, guitar, drum, piano, Each of the multitrack sensing, sound analysis, and sound features 630 of the BPM may be combined to select the play image 640 of the final stream.

However, according to the exemplary embodiment, the split image selection system using the high quality camera according to the exemplary embodiment of the present invention uses only the probability value calculated for the sampled divided image 620 except for the sound feature 630 to generate the final stream. The play video 640 may be selected.

In addition, according to an exemplary embodiment, the plurality of divided images 610 and 620 in FIG. 6 may be an issue scene 541 classified in FIG. 5, and a combination of probability values or sound features 630 among the issue scenes may be the highest. A play video 640 representing a probability value may be selected.

7 illustrates an example of a table for selecting a play video according to an embodiment of the present invention, and FIG. 8 illustrates an example of transmitting a play video according to an embodiment of the present invention.

Referring to FIG. 8, the split image selection system using a high quality camera according to an embodiment of the present invention extracts a plurality of objects A, B, and C from a full image captured by a high quality camera, and extracts a plurality of objects according to the extracted object. A plurality of divided images may be divided, and on-air may be selected by selecting a play image C representing the highest probability value among the divided images.

In the present invention, a table as shown in FIG. 7 may be used to select a play image from among a plurality of divided images.

Referring to FIG. 7, a split image selection system using a high-definition camera according to an embodiment of the present invention sums up scores for each variable of a busking environment for each of a plurality of split images (split images A, B, and C). Play video can be selected.

For example, a split image selection system using a high-definition camera according to an embodiment of the present invention comprises a table of music flow, screen quality / busking lighting, busking space, busking background, and big data. And extract the object for the full image received from the high-definition camera through automatic segmentation, divide the object into a plurality of divided images (A, B, C), and calculate a real-time score for each variable 710 according to each of the plurality of divided images. The final play video (split image C) may be selected based on the total score (probability value) of the calculated result.

Here, the probability value may be calculated from a plurality of divided images sampled based on big data according to an embodiment of the present invention, and the play image may include a multitrack sensing based on a probability value, a musical instrument and a music flow, a sound analysis, and It may be selected by at least one or more sound features of the BPM recognition, or feedback and feedback of the participants.

In addition, as shown in Figure 8, the split image selection system using a high-definition camera according to an embodiment of the present invention can transmit the play video (C) selected in accordance with the probability-based and automatic selection to the real-time chat room through the application In addition, comments and feedback on the play video transmitted from the plurality of participants in the chat room may be received. Accordingly, the split image selection system using the high-definition camera according to an embodiment of the present invention can reselect the play video according to the opinions and feedback of a plurality of participants, and can also enlarge, repeat, or replay the play video. have.

9 is a flowchart illustrating a method of selecting a divided image using a high quality camera according to an embodiment of the present invention.

The split image selection method shown in FIG. 9 is performed by a system for selecting a split image for a busking performance using a high quality camera according to an embodiment of the present invention shown in FIG. 1.

In the method of selecting a split image using a high quality camera according to an exemplary embodiment of the present invention, in step 910, a single full image of a busking performance photographed using the high quality camera is received.

For example, if a busker performs a street performance, the performance producer shoots a busking performance using a single high definition camera, and step 910 is a single pool of busking performances from the performance producer. An image can be received. In this case, the high definition camera may be a UHD (Ultra-HD) 4320P / 4K camera, and step 910 may receive a single 4K full image captured from the high definition camera.

In operation 920, the full image is divided into a plurality of divided images through an object recognition process on the screen of the full image.

For example, in step 920, a person participating in a busking performance on a screen (instantaneous frame) and an object of an instrument used for the busking performance are extracted by using a face (face) or object recognition algorithm. The operation may be divided into a plurality of divided images according to the number of objects.

According to an embodiment, if there is a person who plays only a musical instrument, a singer, or a person who sings a musical instrument in a busking performance, step 920 first extracts the person and each object of the musical instrument, and the person who plays the musical instrument. And, the singer and the person who sings the instrument and the song can be classified separately and divided into divided images for each.

In operation 930, a play image is selected based on a probability of big data according to a busking environment among a plurality of divided images.

In operation 930, a play image may be selected from among a plurality of divided images sampled using big data learned for each music.

Step 930 extracts a sampling of at least one of the bursting environment of music flow, bursting lighting, bursting space, and bursting background from the divided image, compares the extracted result with big data, and displays the plurality of divided images as a general image. Or classifying the image into an issue image (not shown), calculating a probability value for each of the plurality of divided images by applying a machine learning technique to the sampling extracted for the busking environment, and The method may include selecting a play video by combining at least one sound feature of multitrack sensing, sound analysis, and BPM recognition with a probability value.

The apparatus described above may be implemented as a hardware component, a software component, and / or a combination of hardware components and software components. For example, the devices and components described in the embodiments may include, for example, processors, controllers, arithmetic logic units (ALUs), digital signal processors, microcomputers, field programmable arrays (FPAs), It may be implemented using one or more general purpose or special purpose computers, such as a programmable logic unit (PLU), microprocessor, or any other device capable of executing and responding to instructions. The processing device may execute an operating system (OS) and one or more software applications running on the operating system. The processing device may also access, store, manipulate, process, and generate data in response to the execution of the software. For convenience of explanation, one processing device may be described as being used, but one of ordinary skill in the art will appreciate that the processing device includes a plurality of processing elements and / or a plurality of types of processing elements. It can be seen that it may include. For example, the processing device may include a plurality of processors or one processor and one controller. In addition, other processing configurations are possible, such as parallel processors.

The software may include a computer program, code, instructions, or a combination of one or more of the above, and may configure the processing device to operate as desired, or process independently or collectively. You can command the device. Software and / or data may be any type of machine, component, physical device, virtual equipment, computer storage medium or device in order to be interpreted by or to provide instructions or data to the processing device. Or may be permanently or temporarily embodied in a signal wave to be transmitted. The software may be distributed over networked computer systems so that they may be stored or executed in a distributed manner. The software and data may be stored on one or more computer readable recording media.

The method according to the embodiment may be embodied in the form of program instructions that can be executed by various computer means and recorded in a computer readable medium. The computer readable medium may include program instructions, data files, data structures, and the like, alone or in combination. The program instructions recorded on the media may be those specially designed and constructed for the purposes of the embodiments, or they may be of the kind well-known and available to those having skill in the computer software arts. Examples of computer-readable recording media include magnetic media such as hard disks, floppy disks, and magnetic tape, optical media such as CD-ROMs, DVDs, and magnetic disks such as floppy disks. Magneto-optical media, and hardware devices specifically configured to store and execute program instructions, such as ROM, RAM, flash memory, and the like. Examples of program instructions include not only machine code generated by a compiler, but also high-level language code that can be executed by a computer using an interpreter or the like. The hardware device described above may be configured to operate as one or more software modules to perform the operations of the embodiments, and vice versa.

Although the embodiments have been described by the limited embodiments and the drawings as described above, various modifications and variations are possible to those skilled in the art from the above description. For example, the described techniques may be performed in a different order than the described method, and / or components of the described systems, structures, devices, circuits, etc. may be combined or combined in a different form than the described method, or other components. Or, even if replaced or substituted by equivalents, an appropriate result can be achieved.

Therefore, other implementations, other embodiments, and equivalents to the claims are within the scope of the claims that follow.

100, 200: segmentation image selection system using a high quality camera
210: producer
220: host application
230: video production process
240: algorithm
250: participant application
310: high definition camera
510, 620: sampled segmented image
610: segmented split image
630: sound features
640: play video
710: analysis variable

Claims (14)

An image receiver configured to receive a single full image of a busking performance photographed using a high quality camera;
A screen divider for dividing the full image into a plurality of divided images through an object recognition process on the screen of the full image; And
Selecting unit for selecting a play video based on the probability of big data according to the busking environment of the plurality of divided images
Split image selection system using a high-definition camera comprising a. The method of claim 1,
The image receiving unit
Split image selection system using a high-definition camera that receives the single full image of the busking image taken from the single high-definition camera. The method of claim 1,
The screen divider
A split image selection system using a high-definition camera that extracts the object of the person participating in the busking performance and the instrument used in the busking performance, and divides the object into the plurality of divided images according to the number of extracted objects. The method of claim 3,
The screen divider
Split image selection system using a high-definition camera for dividing the extracted person or instrument into the plurality of divided images including the overlapping objects. The method of claim 1,
The selection unit
A split image selection system using a high-definition camera that selects the play image from the plurality of split images sampled using the big data learned for each music. The method of claim 5,
The selection unit
Sampling the at least one of the music flow, the busking illumination, the busking space, and the busking background from the divided image, and extracts a sampling of the bursting environment and compares the extracted result with the big data to generalize the plurality of divided images. Segmentation image selection system using high-definition camera to classify image or issue image. The method of claim 6,
The selection unit
And a high quality camera for calculating a probability value for each of the plurality of divided images by applying a machine learning technique to the extracted sampling for the busking environment. The method of claim 7, wherein
The selection unit
And selecting the play image by combining at least one sound feature of multitrack sensing, sound analysis, and BPM recognition to the probability value. The method of claim 1,
A controller for controlling selection of the play video based on opinions of participants and selection of the play video based on probability among the plurality of divided images
Split image selection system using a high-definition camera further comprising. In the operating method of the system for selecting a divided image for the busking performance using a high-definition camera,
Receiving a single full image of the busking performance photographed using the high quality camera;
Dividing the full image into a plurality of divided images through an object recognition process on the screen of the full image; And
Selecting a play image based on a probability of big data according to a busking environment among the plurality of divided images
Split image selection method using a high-definition camera comprising a. The method of claim 10,
The dividing into the plurality of divided images
The method of selecting a split image using a high-definition camera extracting an object of the person participating in the bursting performance and the instrument used in the bursting performance, and splitting the object into the plurality of divided images according to the number of extracted objects. The method of claim 10,
The step of selecting the play video
The method of selecting a split image using a high-quality camera that selects the play image from among the plurality of split images sampled using the big data learned for each music. The method of claim 12,
The step of selecting the play video
Sampling the at least one of the music flow, the busking illumination, the busking space, and the busking background from the divided image, and extracts a sampling of the bursting environment and compares the extracted result with the big data to generalize the plurality of divided images. Categorizing into an image or an issue image;
Calculating a probability value for each of the plurality of divided images by applying a machine learning technique to the extracted sampling for the busking environment; And
Selecting the play video by combining at least one sound feature of multitrack sensing, sound analysis, and BPM recognition with the probability value;
Split image selection method using a high-definition camera comprising a. A computer program stored in a computer readable recording medium for performing the method of claim 10.

Images