Web Omnidirectional Image Viewer for Field Event Extraction

Abstract

A Web-based image viewer for observing an omnidirectional image of fields is proposed. Conventional monitoring systems are monitoring a narrow area of fields, and their image viewers are not optimized for comparing accumulated images. We implemented two functions of the viewer to compare images and conducted an experiment to evaluate their performance. The experimental results showed that Time-travel mode, which partially shows another image in a circular region on an omnidirectional image, had a high degree of usability for extracting field events from accumulated images. The proposed system can help farmers capture comprehensive, panoramic views of their farm for field management.