High-Frame-Rate Video Analysis for Mobile-Vibration-Based Crop Inspection

Abstract

This paper describes a novel method for mobile vibration-based crop inspection to visualize differences in dynamics between vibrating crops and objects pixel-by-pixel with vision-based FFT analysis. High-frame-rate video analysis was conducted for 1) multiple fans and 2) rice grains and husks on a vibrating sieve, and they were verified by enhancing the differences in dynamics between multiple objects even when their appearances are too similar to distinguish each other in a single image.