Viola-Jones Method of Marker Detection for Scale-Invariant Calculation of Lettuce Leaf Area

Abstract

Leaf area can be used as a growth parameter as such it increases as the stage of lettuce progress. Consideration of scale invariance in estimating the area poses challenging machine vision problems in a smart farm setup. To address this, a marker with known size and components is utilized for the system for normalizing area measurements. This study proposes an automated object detection (marker) using Viola-Jones algorithm that uses Haar features. Based on the result of this study, a high detection rate applied to 40 test samples is obtained by using 30 positive samples and 50 negative samples. The small sample size is compensated by increased number of stages and decreased lower false positive rate for each stage. Future work includes adding training sets and using other methods such as Speeded Up Robust Features (SURF).