Aida Zibaei - Bachelor of Civil engineering, Islamic Azad university of Mashhad, Mashhad, Iran

Today, three-dimensionalization with the help of Lidar cloud points is used in various fields. These include urban extract effects, urban management, computer games, and so on. For this reason, a lot of research has been done in this field. There are many ways to create a ۳D model. The use of image data (single image, stereo images, and multiple images), lidar-based methods, and the use of multisensor data are among the methods of creating a three-dimensional model. In three-dimensional construction, the goal is to increase the accuracy in the model's segmentation, classification, and reconstruction. Therefore, increasing the accuracy in each section will improve the accuracy of ۳D. Inthis regard, there is a need for extensive and detailed research in data classification. Classification refers to the labeling of areas based on the similar properties of objects. Lidar is a relatively new technology for generating spatial data using an active sensor based on a laser beam. Lidar collects object surface information by sending laser beams to the surface of objects and calculating their distance. The purpose of this study is to implement the K-Nearest Neighbors algorithm (KNN) and Support Vector Machine (SVM) to compare the accuracy of urban effects classification based on spatial features, Scattering features, Eigenvalues, etc. By examining the results of the implementationof the above algorithms, the SVM algorithm with a total accuracy of ۷۹.۶۶ is more efficient in classifying Lidar cloud points.

Classification, Machine learning, Support Vector Machine (SVM), K-Nearest Neighbors algorithm (KNN), Cloudy points