Computer Vision Based Methods for Detecting Weeds in Lawns

Abstract

In this paper, two methods for detecting weeds in lawns using computer vision techniques are proposed. The first is based on an assumption about the differences in statistical values between the weed and grass areas in edge images and using Bayes classifier to discriminate them. The second also uses the differences in texture between both areas in edge images but instead applies only simple morphology operators. Correct weed detection rates range from 77.70 to 82.60% for the first method and from 89.83 to 91.11% for the second method. In addition, the proposed methods together with a chemical weeding system as well as a non-chemical weeding system based on pulse high voltage discharge are simulated and the efficiency of the overall systems are evaluated theoretically. With a chemical based system, more than 72% of the weeds can be destroyed with a herbicide reduction rate of 90 to 94% for both methods. For the latter weeding system, killed weed rate varies from 58% to 85%.

Acceleration of Genetic Programming by Hierarchical Structure Learning: A Case Study on Image Recognition Program Synthesis

“…The dataset we used contains 30 images; 20 images are used as a training set and the remaining 10 images as a validation set. The image size is smaller than the original ones used in [27]; its size is 120 × 160 pixels.…”

“…The test problem is the lawn weed detection problem [27]. The goal of GPs is to evolve a program that can accurately segment the area of weeds from lawn backgrounds (Fig.…”

Multi-Objective Genetic Programming with Redundancy-Regulations for Automatic Construction of Image Feature Extractors

“…Four human-designed lawn weed detection methods are considered here. The first is called Bayesian classifier based method (BC) [43]. The second is morphological operation based method (MO) [43].…”

“…Two main performance measures of automatic weed control system are the performance in weed destruction-how weeds can be destroyed (killed), and spraying error-how large area containing no weeds is sprayed. According to the simulated weed control system in [43], after weed detection, the system divides the detected weed image into small blocks of size 30 × 60 pixels (therefore, one image contains 16 × 40 blocks), and it sprays herbicide onto only the blocks containing detected weeds. If a weed is sprayed, they will be counted as a killed (or destroyed) weed.…”

“…Until now, there are a number of methods that have been proposed for detecting weed in lawn fields [3], [17], [37], [39]- [41], [43]. Lawn weed detection problem considered here is a two-class segmentation problem.…”