Radiometric Correction of Terrestrial LiDAR Data for Mapping of Harvest Residues Density

Abstract

ABSTRACT:In precision agriculture detailed geoinformation on plant and soil properties plays an important role. Laser scanning already has been used to describe in-field variations of plant growth in 3D and over time and can serve as valuable complementary topographic data set for remote sensing, such as deriving soil properties from hyperspectral sensors. In this study full-waveform laser scanning data acquired with a Riegl VZ-400 instrument is used to classify 3D point clouds into post-harvest straw residues and bare soil. A workflow for point cloud based classification is presented using radiometric and geometric point features. A radiometric correction is performed by using a range-correction function f(r), which is derived from lab experiments with a reference target of known reflectance. Thereafter, the corrected signal amplitude and local height features are explored with respect to the target classes. The following procedure includes feature calculation, decision tree analysis, point cloud classification and finally result validation using detailed classified reference RGB images. The classification tree separates the classes of harvest residues and bare soil with an accuracy of 96% by using geometric and radiometric features. The LiDAR-derived harvest residue coverage value of 75% lies in accordance with the image-based reference (coverage of 68%). The results indicate the high potential of radiometric features for natural surface classification, particularly in combination with geometric features.

Comparative classification analysis of post-harvest growth detection from terrestrial LiDAR point clouds in precision agriculture

“…Andújar et al (2013) used LiDAR height and reflection measurements for vegetation detection captured at 905 nm wavelength with a detection rate of 77.7% for Sorghum halepense. Similar results are demonstrated by Koenig et al (2013) and Höfle (2014) using height differences in combination with the amplitude in spatial context to distinguish soil and plant objects.…”

“…Post-harvest growth tends to lower amplitude values and elevation differences up to 12 cm compared to ground. The results are comparable to former studies where A mean and the StdZ are chosen for vegetation classification (Höfle, 2014;Koenig et al, 2013;Rutzinger et al, 2008).…”

“…The usage of different approaches on the one hand prevented the usage of classification rules derived from over-fitted modeling of one classifier, and on the other hand substantiated the possibility of classifying the point cloud on the basis of different classification principles. The approaches were selected based on the similar application of those for surface classification (Alexander et al, 2010;Gerke and Xiao, 2014;Pal and Mather, 2003), vegetation detection in airborne (Ducic et al, 2006;Höfle and Hollaus, 2010;Rutzinger et al, 2008;Zlinszky et al, 2012) and in terrestrial LiDAR data (Koenig et al, 2013). Cluster analysis like k-Means has been used in ALS applications such as tree detection (Lindberg et al, 2013;Vauhkonen et al, 2012).…”

TLS Field Data Based Intensity Correction for Forest Environments

“…The latter is based on the fact that for TLS specific applications, the relative range differences between objects and scanner vary much stronger than for ALS. This results in range being one of the most relevant factors for explaining object-independent intensity variations (Höfle, 2014;Koenig et al, 2015;Koenig et al, 2013;Pfeifer et al, 2008). Also, range dependency of TLS in near distance applications does not strictly follow the 1/r 2 law of the radar equation, as is the case for ALS (Höfle, 2014).…”

“…Those are considered to be of major influence to the intensity besides the reflective properties of the object. Höfle (2014), Koenig et al (2015) and Koenig et al (2013) use a less complex approach without calibrated reference targets which is based on the method used before by Höfle and Pfeifer (2007) for ALS data. From the TLS data, as recorded for an agricultural application, they extract large surface areas of homogenous soil to use them as reference for relative calibration of the remaining point cloud.…”