Abstract
Recently, crop growth monitoring has been become urgently needed through the growing season in order to estimate crop yields. Above-ground biomass is one of important crop parameters that can be monitored and assessed by remote sensing technology. In this paper, we aim to estimate above-ground biomass of corn using a non parametric method, k-nearest Neighbors algorithm (kNN) which to extract information about spatial distribution and total carbon stock. A case study area was selected in Shandong Province of China. The satellite image HJ-1 data and Corn biomass were acquired by the synchronous date. Fifteen band ratios were produced and Pearson correlation between spectral properties and corresponding ground data was tested. Based on the analysis, five optimal band ratios were chosen as independent variables. Also, number of k in kNN algorithm was tested; it is from 5 to 15. As a result, the best RMSE is estimated at 118.213 g/m2 when k is 15. Though the number of samples is limited, a model combining HJ-1 and kNN algorithm could be used to create map of above-ground corn biomass.
| Original language | English |
|---|---|
| Publication status | Published - 2014 |
| Event | 35th Asian Conference on Remote Sensing 2014: Sensing for Reintegration of Societies, ACRS 2014 - Nay Pyi Taw, Myanmar Duration: 2014 Oct 27 → 2014 Oct 31 |
Other
| Other | 35th Asian Conference on Remote Sensing 2014: Sensing for Reintegration of Societies, ACRS 2014 |
|---|---|
| Country/Territory | Myanmar |
| City | Nay Pyi Taw |
| Period | 14/10/27 → 14/10/31 |
All Science Journal Classification (ASJC) codes
- Computer Networks and Communications