Three Tests of Agricultural Remote Sensing for Crop Inventory in Eastern Canada: Results, Problems and Prospects

R.A. Ryerson,P. Mosher,V.R. Wallen,N.E. Stewart
DOI: https://doi.org/10.1080/07038992.1979.10854984
IF: 2.6
1979-05-01
Canadian Journal of Remote Sensing
Abstract:LANDSAT digital data and several types of airborne remote sensing imagery have been tested for determining white bean area in Ontario, potato area in New Brunswick, and potato and cropland distribution in Prince Edward Island. White bean area was determined with 85–91% accuracy using 1:104,000 colour infra-red photography and 91–96% accuracy using LANDSAT digital data. Potato area was 94% accurate using 1:112,000 true colour imagery and 85% accurate (for the province) using digital LANDSAT data. LANDSAT data were used in PEI to provide cropland maps — cloud prevented an attempt at a potato inventory. A major contribution of these three tests is in the experience which was gained and which is now shared in this paper. Discussed are the importance of ground based expertise in the crop being studied, test site selection, field data collection and its subsequent organization. Simple methods for handling overlap of spectral signatures, and boundary pixels, accuracy assessment, signature extension, and determining geographic positions on output maps are also detailed.
remote sensing
What problem does this paper attempt to address?