Information processing device for identifying individual animals and activities in using body part extraction

An information processing method performed by an information processing device, comprising:

• an image acquisition step of acquiring a captured analysis target image of a state in which one or more animals are doing activity in a predetermined activity range, the analysis target image comprising a plurality of unit images arranged in a time direction;
• a part extraction step of extracting one or more parts of a body of each of the one or more animals from each of the unit images using a skeleton estimation model that when receiving a unit image, estimates and outputs a skeleton of a body of an animal;
• an individual identification step of identifying the one or more animals as individuals in each of the unit images on the basis of an output obtained by inputting time series of the one or more parts of the body of each of the one or more animals extracted from each of the unit images by the part extraction means to an individual identification model that when receiving time series of one or more parts of a body of an animal, outputs the animal as an individual;
• a specification step of specifying an analysis attribute of the analysis target image;
• a step of selecting, on the basis of the specified analysis attribute of the analysis target image, the skeleton estimation model used in the part extraction step from a plurality of skeleton estimation models and selecting the individual identification model used in the individual identification step from a plurality of individual identification models;
• a generation step of generating a data frame indicating transition of positions of the one or more parts that change with the activity of the one or more animals;
• and
• a activity determination step of, when the transition of the positions of the one or more parts in the data frame satisfies one of previously set one or more conditions, determine activity corresponding to the satisfied condition.

A non-transitory storage medium for causing a computer configured to control an information processing device to perform a control process comprising:

• an image acquisition step of acquiring a captured analysis target image of a state in which one or more animals are doing activity in a predetermined activity range, the analysis target image comprising a plurality of unit images arranged in a time direction;
• a part extraction step of extracting one or more parts of a body of each of the one or more animals from each of the unit images using a skeleton estimation model that when receiving a unit image, estimates and outputs a skeleton of a body of an animal;
• an individual identification step of identifying the one or more animals as individuals in each of the unit images on the basis of an output obtained by inputting time series of the one or more parts of the body of each of the one or more animals extracted from each of the unit images by the part extraction means to an individual identification model that when receiving time series of one or more parts of a body of an animal, outputs the animal as an individual;
• a specification step of specifying an analysis attribute of the analysis target image;
• and
• a model selection step of selecting, on the basis of the specified analysis attribute of the analysis target image, the skeleton estimation model used in the part extraction step from a plurality of skeleton estimation models and selecting the individual identification model used in the individual identification step from a plurality of individual identification models;
• a generation step of generating a data frame indicating transition of positions of the one or more parts that change with the activity of the one or more animals;
• and
• a activity determination step of, when the transition of the positions of the one or more parts in the data frame satisfies one of previously set one or more conditions, determine activity corresponding to the satisfied condition.