ENHANCED FEATURE DETECTION USING LiDAR DATASETS BY NON- TRIVIAL DERIVATION THROUGH ALLOMETRIC AGGREGATION DATA ANALYSIS FOR CHARACTERIZING A DOMINANTLY AGRICULTURAL

derivations This paper provides the application of the Light Detection and Ranging (LiDAR) derived products coupled with orthophotos to derive quality landcover classification using a theoretical approach through principles of data analysis to minimize common problems in image analysis. These problems include misclassification of objects and the non-distinguishable interpretation of pixelated features that results to non-discrimination of class objects due to their closely related spectral resemblance. The research utilizes low density LiDAR point cloud data (2 pts/m2) that accounts for the derived information such as matrices and textures (intensity, nDSM, number of returns, intensity textures, etc.) in the aim of accomplishing the requisites for characteristic selection. Freely accessible vector data and expert based knowledge is integrated in addition to the methodology concerning object-based image analysis. The approach used in this study takes advantage of the principle of allometric relations of two or more observables which are aggregated for each acquisition of data sets to establish a proportionality function for data-partitioning. Non-trivial feature derivations for fitting distributions were employed to construct an ideal hyperplane which we used to separate two or more data sets in distinct regions on a feature space. Accomplishing the distribution computations, allometric relations were analyzed and coupled with the necessary scaling, rotation and transformatio