全部 标题 作者
关键词 摘要

OALib Journal期刊
ISSN: 2333-9721
费用:99美元

查看量下载量

相关文章

更多...
Plant Methods  2011 

Classification of images of wheat, ryegrass and brome grass species at early growth stages using principal component analysis

DOI: 10.1186/1746-4811-7-28

Keywords: Image analysis, image segmentation, principal component analysis, weed detection, plant differentiation

Full-Text   Cite this paper   Add to My Lib

Abstract:

Wheat is the most common agricultural crop in southern Australia and annual ryegrass and brome grass are reportedly the two most common weeds in South Australian wheat fields [1]. These weeds are highly competitive, competing with the crop plants for nutrients at an early stage of growth and producing a large seed bank and subsequently a high number of weeds at emergence. They are host to some cereal diseases and can severely affect wheat yield [1]. Management strategies have not been perfected for weedy grasses in contrast to those used for controlling many broadleaf weeds. These plants are similar in appearance to wheat and require several weeks of growth before distinguishing characteristics and vegetative components fully develop [2]. Nevertheless, the early detection of weed invasions and a quick and coordinated response in order to eradicate them are very important before the weeds become too well established and widespread, making control technically and financially unviable. The weeds that are not detected early may require costly ongoing control efforts [3].The conventional means of manual weed detection is very time consuming, expert-intensive, and costly, even at the early growth stages. On the other hand, early intensive herbicide spraying is not considered an economically and environmentally good option. Therefore, a vision-based and image analysis method was proposed as a cost-effective and site-specific replacement method for weed detection. Digital image analysis has found recent applications in plant biology, plant taxonomy and precision agriculture [4-16]. Perez et al (2000) used a colour RGB camera to detect broadleaf weeds between rows of a cereal crop. Shape analysis was applied for the plants particularly between the rows to detect the weeds. Morphological techniques have been successful to separate broadleaf regions from narrow-leaf plants [17]. A simple approach is to apply a successive erosion process by which the narrow-leaf plants are remo

Full-Text

Contact Us

service@oalib.com

QQ:3279437679

WhatsApp +8615387084133