ABSTRACT: Full-waveform (FWF) lidar system provides both geometric and waveform properties from the entire returned signals for analysis. As it provides more information than the conventional multi echo lidar, the waveform lidar plays an important role in land cover classification as well as object reconstruction. Nowadays, object-based image analysis (OBIA) has been widely applied in multispectral images. The idea of OBIA has been extended to object-based lidar points analysis. The objective of this research is to develop a procedure for object-based lidar points classification using waveform lidar data in a complex scene. There are two main steps in the proposed scheme: (1) point-based segmentation, and (2) object-based classification. Point-based segmentation uses a Euclidean clustering technique and points ’ attributes to merge the neighboring points with similar attributes. After segmentation, an object-based classification rather than point-based classification is performed. Each separated regions after segmentation is a candidate object for classification. An unsupervised Fuzzy c-mean classifier considering the characteristics of curvature, height, echo ratio, echo width, backscattering coefficient and shape information is performed to separate different land covers. The test data is acquired by Rigel Q680i and located in Tainan, Taiwan. The point density is 10 pt/m^2. The experimental result indicates that the proposed method may separate multilayer objects such as tree, building, vehicle, road, and ground. The overall accuracy reached 89 % for waveform features.

This paper describes a method that aims to find all instances of a certain object in Mobile Laser Scanner (MLS) data. In a user-assisted approach, a sample segment of an object is selected, and all similar objects are to be found. By selecting samples from multiple classes, a classification can be performed. Key assumption in this approach is that a one-to-one relationship exists between segments and objects. In this paper the focus is twofold: (1) to explain how to get proper segments, and (2) to describe how to find similar objects. Point attributes that help separating neighbouring objects are presented. These point attributes are used in an attributed connected component algorithm where segments are grown, based on proximity and attribute values. Per component, a feature vector is proposed that consists of two parts. The first is a height histogram, containing information on the height distribution of points within a component. The second contains size and shape information, based on the components ’ bounding box. A simple correlation function is used to find similarities between samples, as selected by a user, and other components. Our approach is tested on a MLS dataset, containing over 300 objects in 13 classes. Detection accuracies heavily depend on the success of the segmentation, and the number of selected samples in combination with the variety of object types in the scene.