In previous studies, signs of hoof disease appear in the curvature of the back of dairy cows and how they walk, so methods for detecting hoof disease by detecting the degree of curvature from images have been studied.However, although earlier detection is an important disease, conventional methods have limited detection to moderate to severe hoof disease.

　Therefore, the research group of Professor Yagi et al. Applied human walking image analysis technology to dairy cows in order to detect hoof disease at an early stage.Specifically, the distance image sensor, which is a camera that can measure the distance to an object, is waterproof and dustproof and installed in the barn of Rakuno Gakuen University, a research cooperation organization.We characterized the state of walking based on the walking images of a large number of dairy cows taken, and detected an individual with hoof disease by machine learning.As a result, hoof disease could be detected with an accuracy of 99% or more in a mild manner.

　Currently, in dairy farming, the number of workers is decreasing and the population is aging, and there is an urgent need to save labor for dairy farmers.It is expected that a "smart barn" will be realized that achieves both labor saving and high quality of products by observing dairy cows in detail that dairy workers cannot see by using artificial intelligence and monitoring technology by video analysis. In the meantime, the results of this research are expected to greatly contribute to the realization of this "smart barn" and to realize a new era of dairy farming where dairy farmers can concentrate on labor saving and high quality products. Will be done.

　The results of this research were presented at the CVIM Study Group of the Information Processing Society of Japan on March 2017 and 3, 10.