مقاله انگلیسی رایگان در مورد تکنیک های پردازش تصویر برای تقسیم بندی گیاهان اراضی – الزویر 2016

1.1. Background and motivation

Weeds are one of the big challenges in agriculture because they appear everywhere randomly, and compete with the plant for resources. As a result of this competition for resources, crop yields suffer. Yield losses depend on factors such as weed species, population density, and relative time of emergence and distribution as well as on the soil type, soil moisture levels, pH and fertility (Papamichail et al., 2002). Numerous researchers have identified a strong link between weed competition and crop yield loss, with a wide range of crop varieties. For example, according to the study by Stall (2009), an annual loss of 146 million pounds of fresh market sweet corn and 18.5 million pounds of sweet corn for processing occurred in the United States from 1975 to 1979 due to weed competition, which corresponds to revenue losses of $13,165,000 and $9,155,000 respectively. Besides, the dry and head weight of crop yield are measured to evaluate losses. Based on a study carried out in 1996/1997 and repeated in 1997/1998 in central Jordan (Qasem, 2009), it was found that the average reduction in shoot dry weight and head yield were 81% and 89% respectively. An effective and efficient weed management system is necessary to minimise yield losses in valuable crops. The critical period for weed control must be taken into account to enhance weed management strategies (Swanton and Weise, 1991), as the duration of co-existence of weed and crop is an important indicator of yield losses due to weed competition (Kropff et al., 1992). Zimdahl (1988, 1993) defined the critical period of weed control (CPWC) as ‘‘a span of time between that period after seeding or emergence when weed competition does not reduce crop yield and the time after which weed competition will no longer reduce crop yield”. A more quantitative definition is as the number of weeks after crop emergence during which a crop must be weedfree in order to prevent yield losses greater than 5% (Hall et al., 1992; Van Acker et al., 1993; Knezevic et al., 1994).