Forest mapping and classification of forest Type using LiDAR data and tree specie identification through image processing based on leaf extraction algorithms

With the use of Light Detection and Ranging (LiDAR) Data, this study focuses on the processing of the LiDAR derived data through different software tools to generate a map that can classify forest types. A 20 × 20 meter plot in the selected forest area was identified in this study for the field validation of the classified leaf type. Leaf recognition is performed using Neural Network in Matlab. The leaf statistics were measured through the prototype developed using leaf extraction algorithms T-test is used for the comparative measurement between the perimeter of the extracted data and the actual perimeter of a sample leaf. The result shows that for the specie, the actual perimeter is statistically the same with the perimeter measured by the developed prototype. The accuracy of classification was calculated as 91.25%. The overall minimum and maximum precision of the prototype is computed to be 90.40% and 99.14%, respectively.