Automatic broccoli detection in RGB-D images

Abstract

Broccoli is a vegetable considered to be of high value worldwide. However, in several countries its production is continually impacted by high labor instability caused by a wide range of economic and political fluctuations and migratory phenomena. For this reason, autonomous and semi-autonomous harvesting alternatives have been sought to facilitate this work, increase production and reduce costs. In this article, we review some proposed strategies for automatic detection in the broccoli harvesting process based on segmentation and 3D image filtering and show the effectiveness of each strategy by comparing the results obtained between them.