Search

Video summarization technologies strive to provide a succinct and thorough description by selecting the most informative frames from the original video. The essential concept of query-based video summarization methods is represented by constructing a video summary related to user query in term of user interest. This paper aims to address the problem of query-based video summarization construction task by adopting a lightweight deep learning model for object detection task in order to evident how concerning a frame or shot is to a given query. Both YOLOv3 and Tiny YOLOv3 deep learning models are used separately in the proposed system to train the networks using images and videos dataset including diverse types of objects such as; (motorcycles, bicycles, cars, buses, and trucks). Subsequence, the most relative frames to a given query are selected and assembled as keyframes using a modified K-mean clustering scheme to provide different interesting summaries from the original video. Based on the experimental results of object detection phase, we have obtained an average object detection accuracy around 93%, 83% based on YOLOv3 and Tiny YOLOv3 deep learning models respectively. Comprehensive experiments were performed to evaluate the proposed video summarization system, which exhibited an efficient summarization rate close to 33% of the original video. Further experiments were conducted using standard UTE dataset to exhibit the competitive performance of the proposed method compared to state of art query-based video summarization methods. [ABSTRACT FROM AUTHOR]

Point cloud voxelization methods are considered as significant topic in 3D computer vision field with variant relevant applications. The massive amount of point cloud data depicted a 3D model are extremely required considerable data storage along with processing time and ordinarily corrupted by outliers and noisy data. In this paper, we propose a new scheme for 3D point cloud clustering based on voxelization without pre-knowledge. Mathematical representation of 3D model is adopted in our framework in order to optimize the point cloud and remove the outliers with preserving the main characteristics of the whole 3D model shape. First, we segment the point cloud data points into uniform rectangular blocks (voxels) based on spatial 3D coordinates of vertices. Subsequence, each voxel is handled as cluster with gravity mass as voxel center and specifies the neighboring points to calculate cluster density. Then, the outliers at each voxel are detected based on geometry features of clusters and adjacency matrix representation. Finally, we reconstruct each voxel by removing the detected outliers. The proposed method is validated on different point cloud models with variant size. The experimental results highlighted that the new clustering-based voxelization method is suitable for real time applications due to its lower processing time. We have got a reduction rate close to 46.8% of the original 3D model size and precision percentage 96 %, which exhibits a higher noise removal rate. [ABSTRACT FROM AUTHOR]

This paper aims to present an effective electromagnetic (EM) modelling approach for circular bundle of single-walled carbon nanotubes (CBSWCNTs), based on the electrical conductivity, relative complex permittivity and linear distribution impedance by applying General Ohm's law for this bundle. The equivalent single conductor material (ESCM) model for personification the CB-SWCNTs is presented in this paper. The main target of this modelling approach is to estimate and investigate the EM properties of CBSWCNTs using common EM engineering tool solver CST (MWS). For this purpose, the CB-SWCNTs and ESCM dipole antennas will be designed and implemented using CST (MWS). Mathematically, the equivalent conductivity model, relative complex permittivity and other parameters of the CB-SWCNTs will be derived in this paper and considered as equivalent material parameters for the ESCM. This modelling technique is expected to provide new avenues for designing different antenna structures. [ABSTRACT FROM AUTHOR]