Non-transitory computer readable recording medium, extraction method, and information processing apparatus

1. A non-transitory computer readable recording medium having stored therein an extraction program that causes a computer to execute a process comprising:

acquiring sensor data and video data of a series of performances that is given by an athlete and includes elements;

extracting positional data of joints of the athlete by analyzing the acquired sensor data;

generating skeleton data and 3D model data of the athlete based on the extracted positional data;

recognizing a posture of the athlete performing each of the elements, based on the generated skeleton data;

extracting from among the elements an element as a first element by determining whether an evaluation item of the element falls within a predetermine range from a first reference for determining success or failure of the element, and an element as a second element by determining whether the evaluation item of the element falls within a predetermined range from a second reference for determining a perfection level of the element, based on the recognized posture and posture success condition stored in a storage unit wherein the posture success condition includes evaluation items;

extracting a first portion of the generated 3D model data corresponding to one of the first element and the second element and a second portion of the acquired video data corresponding to the first portion; and

outputting at least one of the extracted first portion of the generated 3D model data and the extracted second portion of the acquired video data.

2. The non-transitory computer readable recording medium according to claim 1 , wherein

the outputting outputs the extracted first portion of the generated 3D model data that corresponds to 3D data that is generated based on sensor data that is acquired from a viewpoint and a direction from which a portion of the 3D model data of the athlete is viewable, the portion relating to the first reference or the second reference.

3. The non-transitory computer readable recording medium according to claim 1 , wherein

the extracting includes extracting, from among the elements, an element for which a value of a deduction relating to the perfection level is equal to or larger than a predetermined threshold.

4. The non-transitory computer readable recording medium according to claim 1 , wherein

the acquiring acquires sensor data of a plurality of series of performances that is given by different athletes, and

the extracting the first portion extracts first portions respectively for the different athletes and corresponding to the second element for which a difference in deduction values relating to the perfection level exits among the athletes.

5. The non-transitory computer readable recording medium according to claim 1 , further comprising:

when a plurality of second references for determining the perfection level is provided, a certain second reference that is specified from among the plurality of second references is used as the second reference.

6. An extraction method comprising:

acquiring sensor data and video data of a series of performances that is given by an athlete and includes elements, using a processor;

extracting positional data of joints of the athlete by analyzing the acquired sensor data, using the processor;

generating skeleton data and 3D model data of the athlete based on the extracted positional data, using the processor;

recognizing a posture of the athlete performing each of the elements, based on the generated skeleton data, using the processor;

extracting from among the elements an element as a first element by determining whether an evaluation item of the element falls within a predetermine range from a first reference for determining success or failure of the element, and an element as a second element by determining whether the evaluation item of the element falls within a predetermined range from a second reference for determining a perfection level of the element, based on the recognized posture and posture success condition stored in a storage unit wherein the posture success condition includes evaluation items, using the processor;

extracting a first portion of the generated 3D model data corresponding to one of the first element and the second element and a second portion of the acquired video data corresponding to the first portion, using the processor; and

outputting at least one of the extracted first portion of the generated 3D model data and the extracted second portion of the acquired video data, using the processor.

7. An information processing apparatus comprising:

a processor configured to

acquire sensor data and video data of a series of performances that is liven by an athlete and includes elements;

extract positional data of joints of the athlete by analyzing the acquired sensor data;

generate skeleton data and 3D model data of the athlete based on the extracted positional data;

recognize a posture of the athlete performing each of the elements, based on the generated skeleton data;

extract from among the elements an element as a first element by determining whether an evaluation item of the element falls within a predetermine range from a first reference for determining success or failure of the element, and an element as a second element by determining whether the evaluation item of the element falls within a predetermined range from a second reference for determining a perfection level of the element, based on the recognized posture and posture success condition stored in a storage unit wherein the posture success condition includes evaluation items;

a second extracting unit configured to extract a first portion of the generated 3D model data corresponding to one of the first element and the second element and a second portion of the acquired video data corresponding to the first portion; and

an output unit configured to output at least one of the extracted first portion of the generated 3D model data and the extracted second portion of the acquired video data.