Proposal and Evaluation of A Method for Automatically Classifying Images of Agricultural Work and Animals Acquired with Motion Sensor Cameras

We have developed a field monitoring system (http://kansatu.net) using network-connected cameras and sensors installed in apple trees to support the hands-on agriculture curriculum of an elementary school. The system enables to collects and stores a variety of images taken by network-connected cameras equipped with infrared motion sensors. Agricultural works and animals are photographed in these images. The system has been in place seven years since 2011, during which time, there have been years in which more than approximately 50,000 images were collected. However, these data have not been used effectively to learn agricultural works. On the other side, in recent years, damage to agricultural crops by wild animals has been increasing. From these backgrounds, our ultimate goal is to utilize the images captured by motion sensor-equipped network cameras to develop countermeasures to prevent wild animals from damaging agricultural crops and to educate and cultivate future farmers. In this study, we proposed a method that employs image analysis technology and the date and time of image capture to automatically classify images acquired by the motion sensor network cameras by type of agricultural work and animal. We also developed a method for automatically classifying the type of agricultural work. We evaluated the accuracy of the developed method by comparing the results of automated classification with the results of manual classification. The recall of the proposed method exceeded 90% for all three types of agricultural work tested, which was equal to or greater than the classification accuracy achieved with manual classification.