Discrimination of Cabbage Fields for Detecting Clubroot Disease Damage Using Landsat Thematic Mapper Data

[摘要] This paper provides the capabilities of discriminating the growing cabbage fields for detecting clubroot disease damage (a soilborne disease caused by Plasmodiophora brassicae Wor.). Landsat Thematic Mapper (TM) data obtained on July 28, August 13 and 29 1986, and August 8 1990, were used to examine the approach in the entire area of Tsumakoi-mura of Gunma Prefecture. Principal component analysis revealed that the spectral data could be reduced to three eigenvectors (PC1, PC2 and PC3) with a loss of less than 2 percent of the scene variation. They related to the scene brightness (PC1), wetness (PC2) and greenness (PC3) features. Forest and natural vegetation were clearly separable from agricultural crops by the brightness feature. The difference between the cabbage and potato fields was enhanced by the wetness feature due to the dynamic nature of the drying process in potato plants. Harvesting of cabbage affected the reflectance of all bands. The reflectance of visible and middle-infrared bands increased and that of the near-infrared band decreased with the progress of drying in remaining plants. There was a little confusion between the growing and harvested cabbage fields. The optimal choice of TM data for discriminating the growing cabbage is considered to be the period from the beginning to the middle of August.

[学科分类] 农业科学（综合）

[关键词] Cabbage;キャベツ;Classification accuracy;最尤法;Clubroot disease;主成分分析;Landsat Thematic Mapper data;根こぶ病;Linear discriminant analysis;判別分析;Maximum likelihood method;分類精度;Principal component analysis;ランドサットTMデータ