Low altitude multispectral mapping for road defect detection

Pothole’s defect is major damages indicated the road condition visually, and structural defects due to some potential causes. Nowadays, new forms of remote sensing technique were widely used, but less studies in application low altitude multispectral mapping. The images its help better resolution spectral characteristic. Hence, it helps a lot feature classification with proper training sample, classifier band composite. Thus, this study aims extract defective roads by using image Parrot Sequoia flight altitude. This tries detect pothole's existence from combination supervised other than common use which ultimately for agriculture purposes. used Maximum Likelihood, Support vector machine (SVM) Mahalanobis Distance. 15 different probability stacks green, NIR, red edge, as images. comparison performance between types was modeled discussed study. Classifier algorithm maximum likelihood gives lowest error 0.108m² NIR + edge band. SVM 0.427m² green While distance -0.082m² Averagely, 0.299m² all bands used. Keywords: Road, pothole, UAV, multispectral, UAV detection