Vegetation Mapping of a Tomato Crop using Multilayer Perceptron (MLP) Neural Network in Images Acquired by Remote Sensing from a UAV

Remote sensing from an Unmanned Aerial Vehicle (UAV), also known as Low Altitude Remote Sensing provides interesting options for applications in agriculture. Vegetation mapping is an important application in remote sensing applications. In this work vegetation mapping is carried out in a tomato crop. Aerial imagery of tomato crop is acquired by a Quadcoptor UAV with an optical sensor as the payload. The optical sensor is the camera module of a Raspberry PI single board Single Board Computer (SBC). Vegetation mapping of the tomato crop is carried out by segmentation of the tomato crop images using the proposed MLP-SEG method. Performance of MLP-SEG method is compared with a Support Vector Machine (SVM) based method SVM-SEG. Confusion matrix parameters are used to analyse the performance of the proposed method. The results indicate that MLP-SEG performance is comparable to SVM-SEG. General Terms Vegetation mapping, Unmanned Aerial Vehicle (UAV), Low Altitude Remote Sensing (LARS),