Journal of Computer and Robotics
Volume:11 Issue: 2, Summer and Autumn 2018

The most commonly used method for the production of thin films is based on deposition of atoms or molecules onto a solid surface. One of the suitable method is to produce high quality metallic, semiconductor and organic thin film is Ionized cluster beam deposition (ICBD), which are used in electronic, robotic, optical, optoelectronic devices. Many important factors such as cluster size, cluster energy, impact angle and substrate temperature have important effects on the quality of final thin film such as cluster implanted atoms, substrate sputtering atoms and surface roughness. In this paper, molecular dynamics (MD) simulation of nano-Si cluster impact on Si(100) substrate surface has been carried out for energies of 1-5 eV/atom. The 3-body Stillinger-Weber potential (SW) was used in this simulation. Si cluster sizes of 30, 70, and 160 atoms were deposited on a Si (100) substrate whose temperatures were set around 300 K. Our results illustrate that the maximum substrate temperature, heat transferred time, the cluster implantation and sputtering atoms from the surface increase with increasing the cluster size and energy of the clusters. We found that small nano-clusters with high kinetic energy can produce flatter surface.

Grid computing environments include heterogeneous resources shared by a large number of computers to handle the data and process intensive applications. In these environments, the required resources must be accessible for Grid applications on demand, which makes the resource discovery as a critical service. In recent years, various techniques are proposed to index and discover the Grid resources. The response time and message load during the search process could highly affect the efficiency of resource discovery. This paper proposes a new technique in order to forward the queries based on the resource types which are accessible through each branch in hierarchical Grid resource discovery approaches. To evaluate the proposed technique, it is simulated in GridSim. The experimental results showed reducing the response time and message load during the search process especially when the Grid environment contains a large number of nodes.

Because of the low accuracy of estimation and uncertainty of the techniques used in the past to Software Cost Estimation (SCE), software producers face a high risk in practice with regards to software projects and they often fail in such projects. Thus, SCE as a complex issue in software engineering requires new solutions, and researchers make an effort to make use of Meta-heuristic algorithms to solve this complicated and sensitive issue. In this paper, we propose a new method by improving Genetic Algorithm (GA) with Bat Algorithm (BA), considering the effect of qualitative factors and false variables in the relations concerning the total estimation of the cost. The proposed method was investigated and assessed on four various datasets based on seven criteria. The experimental results indicate that the proposed method mainly improves accuracy in the SCE and it reduced errors' value in comparison with other models. In the results obtained, Mean Magnitude of Relative Error (MMRE) on NASA60, NASA63, NASA93, and KEMERER is 17.91, 34.80, 41.97, and 95.86, respectively. In addition, the experimental results on datasets show that the proposed method significantly outperforms GA and BA and also many other recent SCE methods.

Multiple description (MD) coding has evolved as a promising technique for promoting error resiliency of multimedia system in real-time application programs over error-prone communicational channels. Although multiple description lattice vector quantization (MDCLVQ) is an efficient method for transmitting reliable data in the context of potential error channels, this method doesn’t consider discreteness of network so that losing all descriptions is highly possible. It means all videos may be removed. In this study, we have implemented scheme of MDCLVQ in real-time environment of network, in a method that, raw video (i.e. video with no standard encoding (like MPEG)) is transmitted through independent packets inside of network. This technique leads in low or close to zero loss of all packets. Our purpose is to increase error resiliency and reliable data transmission in error-prone channels. The technique has been tested on some videos sources of Akiyo, Carphone, Foreman and Miss-America. The experimental results indicate that quality of reconstructed videos are substantially improved in terms of central and side PSNR.

Scene change detection plays an important role in a number of video applications, including video indexing, searching, browsing, semantic features extraction, and, in general, pre-processing and post-processing operations. Several scene change detection methods have been proposed in different coding standards. Most of them use fixed thresholds for the similarity metrics to determine if there was a change or not. These thresholds are obtained empirically or they must be calculated before the scene change detection after the whole sequence is obtained. Efficiency of scene change detectors decreases considerably for videos with high scene complexity and variation. In this paper, we propose a novel scene change detection algorithm in the HEVC compressed domain. In the proposed method, we have developed an efficient method based on the analysis of the Transform Units distribution in HEVC standard. In order to enhance the accuracy of detecting the scene changes, we have also defined an automated, dynamic threshold model which can efficiently trace scene changes. The experimental results on UHD videos indicate a higher performance with significantly improved accuracy combined with minimum complexity.

Distributed mutual exclusion is a fundamental problem of distributed systems that coordinates the access to critical shared resources. It concerns with how the various distributed processes access to the shared resources in a mutually exclusive manner. This paper presents fully distributed improved token based mutual exclusion algorithm for distributed system. In this algorithm, a process which has owing token, could enter to its critical section. The processes communicate to each other in an asynchronous message passing manner. We assume the distributed processes are organized in a wraparound two dimensional array. Also, the communication graph of the network is supposed to be a complete graph. The proposed algorithm uses three types of messages, namely ReqMsg, InfoMsg and RelMsg. Beside token-holding node, there are some nodes, we call them informed-nodes, which can know token-holding node and transmit request message to it directly. The number of messages, which are exchanged per each critical section entrance, is a key parameter to avoid posing additional overhead to the distributed system. In this paper, we obtain to  messages per critical section access where N is the number of nodes in the system. The proposed algorithm outperforms other token based algorithms whilst fairness is kept and the proposed algorithm is starvation free.

The intelligence of a mobile robot is highly dependent on its vision. The main objective of an intelligent mobile robot is in its ability to the online image processing, object detection, and especially visual tracking which is a complex task in stochastic environments. Tracking algorithms suffer from sequence challenges such as illumination variation, occlusion, and background clutter, so an accurate tracker should employ the appropriate visual features to identify target. In this paper, we propose using the histogram of oriented gradient (HOG), as an important descriptor. The descriptor simulates the performance of the complex cells in the primary visual cortex (V1) and it has low sensitivity to the illumination changes. In the proposed method, firstly, an object model is generated by training the HOG of multi first frames via an SVM classifier. Then, in order to track a new frame, the HOG descriptors are extracted from the surrounding areas of the target in the previous frame and convolved with the object model. Finally, the location with the highest score is defined as the target. The experimental results demonstrate the proposed method has significant performance compare to the state-of-the-art methods. Furthermore, we apply our algorithm to the mobile robot built by the robotics team to ensure its performance in a real environment.

One of the most important artificial immune algorithms is negative selection algorithm, which is an anomaly detection and pattern recognition technique; however, recent research has shown the successful application of this algorithm in data classification. Most of the negative selection methods consider deterministic boundaries to distinguish between self and non-self-spaces. In this paper, two negative selection based algorithms are proposed for two-class and multi-class classification problems; using a Gaussian mixture model which is fitted on normal space to create a flexible boundary between self and non-self-spaces, by determining the dynamic subsets of effective detectors to solve the problem of data classification. Initialization of effective parameters such as the detection threshold, the maximum number of detectors etc. for each dataset, is one of the challenges in negative selection based classification algorithms, which affects the precision and accuracy of the classification; therefore, an adaptive and optimal calculation of these parameters is necessary. To overcome this problem, the particle swarm optimization algorithm has been used to properly set the parameters of the proposed methods. The experimental results showed that using a Gaussian mixture model and dynamic adjustment of parameters such as optimum number of Gaussian components according to the shape of the boundaries, creation of appropriate number of detectors, and also automatic adjustment of the training and testing thresholds, using particle swarm optimization algorithm as well as utilization of a combinatorial objective function has led to a better classification with fewer detectors. The proposed algorithms showed competitive performance compared with some of the existing classification algorithms, including several immune-inspired models, especially negative selection ones, and other traditional classification methods.