Marco Piragnolo

Towards Surveying with a Smartphone

Photogrammetry is one of the most used techniques for monitoring and surveying. It is widely used in several applications and in different working conditions. Accuracy of photogrammetry reconstruction methods may change depending on the working conditions (e.g. the number of acquired images, lighting conditions, baselines between images), and it is strictly related to the success of the solution of the Structure from Motion problem. Despite its widely spread use and the ever growing improvements to the reconstruction technique, photogrammetry still does not reach the same level of reliability of laser scanning surveying techniques: significant issues may occur in photogrammetric reconstructions when in presence of lighting problems or when the object of interest is not sufficiently textured. However, it relies on the use of much cheaper tools with respect to laser scanning techniques and surveying is usually much faster. This paper aims at showing the potential improvement that can be obtained by introducing information provided by the navigation system in the 3D reconstruction algorithm: the goal is that of making the solution algorithm of the Structure from Motion problem more reliable and accurate. As a side effect, faster reconstruction is typically achieved. The technique is validated on a building using images and navigation information got from a standard smartphone.

Open source R for applying machine learning to RPAS remote sensing images

The increase in the number of remote sensing platforms, ranging from satellites to close-range Remotely Piloted Aircraft System (RPAS), is leading to a growing demand for new image processing and classification tools. This article presents a comparison of the Random Forest (RF) and Support Vector Machine (SVM) machine-learning algorithms for extracting land-use classes in RPAS-derived orthomosaic using open source R packages.The camera used in this work captures the reflectance of the Red, Blue, Green and Near Infrared channels of a target. The full dataset is therefore a 4-channel raster image. The classification performance of the two methods is tested at varying sizes of training sets. The SVM and RF are evaluated using Kappa index, classification accuracy and classification error as accuracy metrics. The training sets are randomly obtained as subset of 2 to 20% of the total number of raster cells, with stratified sampling according to the land-use classes. Ten runs are done for each training set to calculate the variance in results. The control dataset consists of an independent classification obtained by photointerpretation. The validation is carried out(i) using the K-Fold cross validation, (ii) using the pixels from the validation test set, and (iii) using the pixels from the full test set.Validation with K-fold and with the validation dataset show SVM give better results, but RF prove to be more performing when training size is larger. Classification error and classification accuracy follow the trend of Kappa index.