Journal of Information Science and Engineering
- OpenAccess
社團法人中華民國計算語言學學會,正常發行
選擇卷期
- 期刊
- OpenAccess
Material design is an important task in game and animation industry. It often requires artist's great efforts to create an elaborate material. Although several successful material editing systems already exist, it is still hard for artists to design a material on their own since these systems usually have neither an intuitive and sufficiently sophisticated material representation nor an artist-friendly editing interface. In this paper, we propose a simple material creation flow and develop a material design system based on an intuitive material representation. With the system, an artist can bring their creativity into play on designing materials that are realistic or imaginary.
- 期刊
- OpenAccess
Reversible data hiding (RDH) can extract secret messages and restore the original image without distortion. The reversibility benefits many practical applications such as medical image processing and multimedia archive management. Because of high image quality, histogram modification is applied to RDH in many literatures. In this paper, two-dimensional histogram and prediction-error expansion are integrated with a well-designed difference-pair mapping (DPM) to improve embedding capacity. According to the simulation results, the proposed RDH scheme outperforms existing approaches, which are also based on two-dimensional histogram. On the average, image quality is improved by 3dB with the same embedding capacity. Maintaining at the same image quality, embedding capacity can be improved by approximately 30,000 bits.
- 期刊
- OpenAccess
Elastic optical networks (EONs) are considered as very promising architectures for future optical transport networks, since they efficiently use the spectrum resources and provide high bandwidth scalability and granularity. In this paper, the virtual topology design (VTD) problem on EON is considered. Given the physical network and the traffic demand matrix, the goal of VTD is to find the routing paths and the allocated frequency slots of the demand so that the objective cost can be minimized. In this paper, multiplepath routing is allowed and delay-variation between lightpaths for same demand is considered. In this paper, an integer linear programming (ILP) model is used to define the VTD problem on EONs. Moreover, a genetic algorithm (GA) and two heuristic algorithms are proposed to solve this problem. Simulations show that the proposed algorithms can achieve good results.
- 期刊
- OpenAccess
This paper proposes a virtual touchpad human computer interaction (HCI) based on the fingertip hand gesture recognition (FHGR) method for smartphone with depth camera. The virtual touchpad HCI requires less computing power and memory resource than the conventional hand gesture recognition (HGR) methods. It consists of three modules: (1) open hand identification, (2) active fingertip detection and tracking, and (3) fingertip hand gesture recognition (FHGR). First, the handshape region is segmented and divided into three component regions: arm, palm, and fingers. Second, we locate the fingertips and then detect the active fingertips. Finally, the trajectory of every active fingertip is analyzed for FHGR. The two major contributions of this paper are (1) development of a real-time FHGR method with high recognition accuracy and (2) implementation of a virtual touchpad HCI for the smart phones.
- 期刊
- OpenAccess
Content-Based Image Retrieval (CBIR) systems are widely used for local as well as for remote applications such as telemedicine, satellite image transmission and image search engines. The existing CBIR systems suffer from the limitations of storage space, data security and bandwidth requirement. To overcome these problems, a new method termed as CBIR-PE which makes use of prediction errors instead of actual images for storage, transmission and retrieval is presented. Identical artificial neural networks (ANNs) are employed both at the server and client sides to carry out the prediction. At the server side, only the error database comprising the difference between the original and the predicted pixel values is used instead of the actual image database. The prediction errors of the query image are matched with those in the server database to retrieve similar prediction error patterns. These errors are then combined with the predicted values available at the client ANN to reconstruct the actual images. Since only the prediction errors are employed, the proposed method is able to solve the problems of storage space, data security and bandwidth requirement. The proposed method is implemented in combination with a clustering technique called WBCT-FCM which makes use of wavelet based contourlet transform (WBCT) and fuzzy c-means (FCM) clustering algorithm. The performances of the proposed WBCT-FCM and CBIR-PE are evaluated using COREL- 1k database. The experimental results show that the proposed methods achieve better results with respect to clustering and retrieval accuracies compared to the existing methods.
- 期刊
- OpenAccess
Elastic Optical Networks (EONs) are considered as very promising architectures for future optical transport networks, since they efficiently use the spectrum resources and provide high bandwidth scalability and granularity. In this paper, the multicast routing problem in EON is considered. A fragmentation-aware multicast routing algorithm, named as Weighted Dynamic Fragmentation-aware Multicast Routing Algorithm (WDFMRA), is proposed to solve this problem. Simulations show that the fragmentation and blocking ratio of the proposed algorithm can achieve good results.
- 期刊
- OpenAccess
A continuous range query over moving objects continually retrieves the moving objects that are currently within a given query region of interest. Most existing approaches assume that moving objects continually communicate with the server to report their current locations and the server updates the results of queries continuously. However, this assumption degrades the system performance because the communication cost and the server workload increase when the number of moving objects and queries becomes huge. The QR-tree is a query indexing structure, which helps the server cooperate with the moving objects efficiently by utilizing the available computational resources of the moving objects to improve the overall system performance. In this paper, we propose a variant of the QR-tree, namely, the QR*-tree, which helps reduce (i) the amount of location- update stream generated from moving object and (ii) the server work load for query evaluation. Through a series of comprehensive simulations, we verify the efficiency of the QR*-tree in terms of the wireless communication cost and the server workload.
- 期刊
- OpenAccess
A directed hypergraph has hyperedges that connect multiple source nodes and multiple target nodes. These hyperedges are useful to model relationships between multiple entities, which often occur in the biological network. This paper focuses on the problem to find a minimum cardinality source set shortly, a Minimum Source Set (MinSS) that can reach a set of given target nodes. However, this problem is NP-hard, and hence requires a large amount of computation time even for a problem with a small size. We propose the MSS index where for each node the set of precomputed all minimal source sets is maintained. By constructing an MSS index in advance, a MinSS for any node can be obtained in a very short time. To avoid too much large size of the MSS index for a very large hypergraph, we propose a reduction method using the source set proxy. We also describe how to maintain the MSS index for insertion and deletion of nodes and hyperedges. We show through performance experiments that our proposed method can be a viable solution for the MinSS problem in the sense that the MSS index can handle the problems with reasonably large sizes in acceptable amounts of time.
- 期刊
- OpenAccess
Due to the large bandwidth offered by the optical fiber, its implementation in local area network (LAN) where the traffic is typically bursty could be considered as the strongest candidates for the future high speed optical networks if coding technique is used. This paper presents a new code family named, diagonal permutation shifting (DPS) for spectral amplitude coding-optical code division multiple access (SAC-OCDMA) systems with fixed in-phase cross correlation (CC) and short code length. DPS code derived from well-known prime codes and some matrix operations. The results reveal that the DPS code acts effectively in removing multiple access interface (MAI) and mitigate phase induced intensity noise (PIIN) impact with an adequate signal to interference plus-noise ratio (SINR) and low bit error rate (BER). It has been shown that the DPS code family can suppress intensity noise effectively and improve the system performance significantly.
- 期刊
- OpenAccess
Keyloggers have been studied for many years, but they still pose a severe threat to information security. Keyloggers can record highly sensitive information, and then transfer it to remote attackers. Previous solutions suffer from limitations in that: (1) Most methods focus on user-level keylogger detection; (2) Some methods need to modify OS kernels; (3) Most methods can be bypassed when the OS kernel is compromised. In this paper, we present LAKEED, an online defense against the kernel-level keylogger by utilizing the hardware assisted virtualization technology. Our system is compatible with the commodity operating system, and it can protect the running OS transparently. The basic idea of our approach is to isolate the target kernel extension that may contain the keylogger from keyboard drivers' execution environment and then monitor their potential interactions. By comparing the runtime information with the execution baseline that is obtained by the offline analysis, the keylogger can be identified. The evaluation shows that LAKEED can defeat kernel-level keyloggers effectively with low performance overhead.