The Cluster analysis is a major technique for statistical analysis, machine learning, pattern recognition, data mining, image analysis and bioinformatics. K-means algorithm is one of the most important clustering algorithms. However, the k-means algorithm needs a large amount of computational time for handling large data sets. In this paper, we developed more efficient clustering algorithm to overcome this deficiency named Fast Balanced k-means (FBK-means). This algorithm is not only yields the best clustering results as in the k-means algorithm but also requires less computational time. The algorithm is working well in the case of balanced data.

The Cluster analysis is a major technique for statistical analysis, machine learning, pattern recognition, data mining, image analysis and bioinformatics. K-means algorithm is one of the most important clustering algorithms. However, the k-means algorithm needs a large amount of computational time for handling large data sets. In this paper, we developed more efficient clustering algorithm to overcome this deficiency named Fast Balanced k-means (FBK-means). This algorithm is not only yields the best clustering results as in the k-means algorithm but also requires less computational time. The algorithm is working well in the case of balanced data.

The challenge of large-scale content-based image retrieval (CBIR) has been recently addressed by many promising approaches. In this work, a new approach that jointly optimizes the search precision and time for large-scale CBIR is presented. This is achieved using binary local image descriptors, such as BRIEF or BRISK, along with binary hashing methods, such as Locality-Sensitive Hashing and Spherical Hashing (SH). The proposed approach, named Multi-Bin Search, improves the retrieval precision of binary hashing methods through computing, storing and indexing the nearest neighbor bins for each bin generated from a binary hashing method. Then, the search process does not only search the targeted bin, but also it searches the nearest neighbor bins. To efficiently search inside targeted bins, a fast exhaustive-search equivalent algorithm, inspired by Norm Ordered Matching, has been used. Also, a result reranking step that increases the retrieval precision is introduced, but with a slight increase in search time. Experimental evaluations over famous benchmarking datasets (such as the University of Kentucky Benchmarking, the INRIA Holidays, and the MIRFLICKR-1M) show that the proposed approach highly improves the retrieval precision of the state-of-art binary hashing methods.

The challenge of large-scale content-based image retrieval (CBIR) has been recently addressed by many promising approaches. In this work, a new approach that jointly optimizes the search precision and time for large-scale CBIR is presented. This is achieved using binary local image descriptors, such as BRIEF or BRISK, along with binary hashing methods, such as Locality-Sensitive Hashing and Spherical Hashing (SH). The proposed approach, named Multi-Bin Search, improves the retrieval precision of binary hashing methods through computing, storing and indexing the nearest neighbor bins for each bin generated from a binary hashing method. Then, the search process does not only search the targeted bin, but also it searches the nearest neighbor bins. To efficiently search inside targeted bins, a fast exhaustive-search equivalent algorithm, inspired by Norm Ordered Matching, has been used. Also, a result reranking step that increases the retrieval precision is introduced, but with a slight increase in search time. Experimental evaluations over famous benchmarking datasets (such as the University of Kentucky Benchmarking, the INRIA Holidays, and the MIRFLICKR-1M) show that the proposed approach highly improves the retrieval precision of the state-of-art binary hashing methods.

The challenge of large-scale content-based image retrieval (CBIR) has been recently addressed by many promising approaches. In this work, a new approach that jointly optimizes the search precision and time for large-scale CBIR is presented. This is achieved using binary local image descriptors, such as BRIEF or BRISK, along with binary hashing methods, such as Locality-Sensitive Hashing and Spherical Hashing (SH). The proposed approach, named Multi-Bin Search, improves the retrieval precision of binary hashing methods through computing, storing and indexing the nearest neighbor bins for each bin generated from a binary hashing method. Then, the search process does not only search the targeted bin, but also it searches the nearest neighbor bins. To efficiently search inside targeted bins, a fast exhaustive-search equivalent algorithm, inspired by Norm Ordered Matching, has been used. Also, a result reranking step that increases the retrieval precision is introduced, but with a slight increase in search time. Experimental evaluations over famous benchmarking datasets (such as the University of Kentucky Benchmarking, the INRIA Holidays, and the MIRFLICKR-1M) show that the proposed approach highly improves the retrieval precision of the state-of-art binary hashing methods.

The challenge of large-scale image retrieval has been recently addressed by many promising approaches. In this work, we propose a new approach that jointly optimizes the search accuracy and time by using binary local image descriptors, such as BRIEF and BRISK, and binary hashing methods, such as Locality Sensitive Hashing (LSH) and Spherical Hashing. We propose a Multi-bin search method that highly improves the retrieval precision of binary hashing methods. Also, we introduce a reranking scheme that increases the retrieval precision, but with a slight increase in search time. Evaluations on the University of Kentucky Benchmark (UKB) dataset show that the proposed approach greatly improves the retrieval precision of recent binary hashing approaches.

The challenge of large-scale image retrieval has been recently addressed by many promising approaches. In this work, we propose a new approach that jointly optimizes the search accuracy and time by using binary local image descriptors, such as BRIEF and BRISK, and binary hashing methods, such as Locality Sensitive Hashing (LSH) and Spherical Hashing. We propose a Multi-bin search method that highly improves the retrieval precision of binary hashing methods. Also, we introduce a reranking scheme that increases the retrieval precision, but with a slight increase in search time. Evaluations on the University of Kentucky Benchmark (UKB) dataset show that the proposed approach greatly improves the retrieval precision of recent binary hashing approaches.

The challenge of large-scale image retrieval has been recently addressed by many promising approaches. In this work, we propose a new approach that jointly optimizes the search accuracy and time by using binary local image descriptors, such as BRIEF and BRISK, and binary hashing methods, such as Locality Sensitive Hashing (LSH) and Spherical Hashing. We propose a Multi-bin search method that highly improves the retrieval precision of binary hashing methods. Also, we introduce a reranking scheme that increases the retrieval precision, but with a slight increase in search time. Evaluations on the University of Kentucky Benchmark (UKB) dataset show that the proposed approach greatly improves the retrieval precision of recent binary hashing approaches.

Augmented Reality ‘‘AR’’ is a promising paradigm that can offer users with real-time, high- quality visualization of a wide variety of information. In AR, virtual objects are added to the real-world view in real time. The AR technology can offer a very realistic environment for enhancing drivers’ performance on the road and testing drivers’ ability to react to different road design and traffic operations scenarios. This can be achieved by adding virtual objects (people, vehicles, hazards, and other objects) to the normal view while driving an actual vehicle in a real environment. This paper explores a new Augmented Reality Vehicle ‘‘ARV’’ system and attempts to apply this new concept to a selected traffic engineering application namely the left-turn maneuver at two-way stop-controlled ‘‘TWSC’’ intersec- tion. This TWSC intersection experiment, in addition to testing the feasibility of the appli- cation, tries to quantify the size of gaps accepted by different driver’s characteristics (age and gender). The ARV system can be installed in any vehicle where the driver can see the surrounding environment through a Head Mounted Display ‘‘HMD’’ and virtual objects are generated through a computer and added to the scene. These different environments are generated using a well defined set of scenarios. The results from this study supported the feasibility and validity of the proposed ARV system and they showed promise for this system to be used in the field-testing for the safety and operation aspects of transportation research. Results of the left-turn maneuver study revealed that participants accepted gaps in the range of 4.0–9.0 s. This finding implies that all gaps below 4 s are rejected and all gaps above 9 s are likely to be accepted. The mean value of the left-turn time was 4.67 s which is a little bit higher than reported values in the literature (4.0–4.3 s). Older drivers were found to select larger gaps to make left turns than younger drivers. The conservative driving attitude of older drivers indicates the potential presence of reduced driving ability of elderly. Drivers’ characteristics (age and gender) did not significantly affect the left-turn time. Based on the survey questions that were handed to participants, most participants indicated good level of comfort with none or small level of risk while driving the vehicle with the ARV system. None of the participants felt any kind of motion sickness and the participants’ answers indicated a good visibility and realism of the scene with overall good system fidelity.

Abstract—Applying new technologies to traffic engineering studies has become more urgent due to the high cost and risk associated with ordinary in-the-field testing. Augmented reality (AR) is one of those technologies, in which virtual (computer- generated) objects are added to the real scene in a way that the user cannot distinguish between real and virtual objects in the final scene. Adding virtual objects (people, vehicles, hazards, and other objects) to the normal view can provide a safe realistic environ- ment for testing driving performance under different scenarios. This paper presents two systems, i.e., AR vehicle (ARV) and offline AR simulator (OARSim) systems, and uses them to study the left- turn driving behavior at an unsignalized intersection for drivers with different characteristics. Two experiments were performed: one using the ARV system installed in a vehicle and another using the OARSim system installed in the laboratory. Quantitative mea- surements of left-turn drivers’ behaviors were recorded. There was no significant gender effect on all measured parameters in both experiments. Older drivers selected larger gaps and used smaller acceleration rates to turn left than younger drivers in both experiments. The conservative driving attitude of older drivers indicates the potential presence of reduced driving ability of the elderly. While left-turn times using the ARV system were not sig- nificantly affected by drivers’ age, older drivers took longer time to complete the left-turn maneuver than younger drivers using the OARSim did. Results from this study supported the feasibility and validity of the proposed systems and showed promise for these systems to be used as surrogates to in-the-field testing for safety and operation aspects of transportation research.