Article ID Journal Published Year Pages File Type
6540194 Computers and Electronics in Agriculture 2016 8 Pages PDF
Abstract
Early disease detection in greenhouses is an important component of an integrated management strategy. An automated process can be more accurate and cost effective. Automated disease detection of plants in the field is complex due to an uncontrolled environment, and the scale at which detection must be done. This paper presents a powdery mildew detection machine vision system for use in greenhouses. The system uses augmented lighting to remove clutter and reduce the search space, and a Hough forest to detect early powdery mildew. Field testing on greenhouse tomato plants has showed detection rates of 85%.
Related Topics
Physical Sciences and Engineering Computer Science Computer Science Applications
Authors
, ,