High resolution remote sensing is a valuable tool for quantifying the distribution and density of trees with applications ranging from forest inventory, mapping urban parklands to understanding impacts on soil nutrient and carbon dynamics in farming land. The present study aims to compare the accuracy of different remote sensing techniques for delineating the tree cover in 50 cm resolution WorldView2 imagery of farmland. An image of farmland comprising pastures, remnant vegetation and woodland was initially classified into six classes, namely tree cover, bare soil, rock outcrop, natural pasture, degraded pasture and water body using different techniques. Pixel based classification based on all four available wavebands, were tested and an overall classification accuracy of 96.8% and 72.9 % were achieved for supervised and unsupervised techniques. Object based segmentation and subsequent classification yielded an improved overall classification accuracy of 98.3%. Addition of a fifth NDVI layer to the available wavebands did improve the accuracy but not significantly (98.1%, approx 1.3%). In addition to the improvements in overall classification accuracy, a visual inspections of results from the different methods indicated the object based method to yield a more 'realistic' result, avoiding the 'salt and pepper' effects apparent in the pixel-based methods. Overall, object based classification hence is considered more suitable for tree cover extraction from high resolution images.

AN ANALYSIS OF time series Landsat imagery acquired over the Bundaberg region between 2010 and 2015 identified variations in annual crop vigour trends, as determined by greenness normalised difference vegetation index (GNDVI). On average, early to mid-April was identified as the crucial period where crops achieved their maximum vigour and as such indicated when single image captures should be acquired for future regional yield forecasting. Additionally, the regional crop GNDVI averaged from Landsat images between February to April, produced a higher coefficient of determination to final yield (R2 = 0.91) than the average crop GNDVI value from a single mid-season SPOT5 image capture (R2 = 0.52). This result indicates that the time series method may be more appropriate for future regional yield forecasting. For improved prediction accuracies at the individual crop level, a univariate model using only crop GNDVI values (SPOT5) and corresponding yield (t/ha) produced a higher prediction accuracy for the 2014 Bundaberg harvest than a multivariate model that included additional historic spectral and crop attribute data. For Condong, a multivariate model improved the prediction accuracy of individual crops harvested in 2014 by 41.8% for one-year-old cane (Y1), and 46.2% for two-year-old cane (Y2). For the non-invasive measure of foliar nitrogen (N%), the specific wavelengths 615 nm, 737 nm and 933 nm (Airborne hyperspectral), and 634 nm, 750 nm and 880 nm (ground based field spectroscopy) were found to be the most significant. These results were supported by satellite imagery (Worldview-2 and Worldview-3) acquired over three replicated field trials in Mackay (2014 and 2015) and Tully (2015), where the vegetation index (VI) REN2NDVIWV, a ratio of the rededge band (705-745 nm) and the Near-IR2 band (860-1040 nm), produced a higher correlation to nitrogen concentration (%) than NDVI.