Scientia Iranica
Volume:22 Issue: 6, 2015

In this paper, a method using an evolutionary algorithm to automatically set-up the color-feature model of an omnidirectional vision system will be introduced. The mentioned method, in addition to avoiding the issue of over-reliance on lighting conditions when the soccer robot is performing image processing, can also very e ectively speed up the parameter setup procedure of the robot vision system. Hence, when the robot is moving in the soccer eld, it can nish target object detection and self-localization in real time. In order to verify the e ectiveness of the mentioned method, tests have been conducted under di erent bad lighting conditions, and the experimental results show that the soccer robot can always set up the parameters of the vision system. It can also set up the color-feature model that is applicable to the operational environment at that moment and detect target objects such as goals and the eld. Meanwhile, through relative location between detected target objects and the robot, self-localization and path planning can be nished.

General boosting algorithms for face detection use rectangular features. To obtain a better performance, it needs more training samples and may generate an unpredictable number of features. Besides using pixel values, which are easily a ected by illumination, to calculate the rectangular features, it usually needs to preprocess the data before calculating the values of the features. Such an approach may increase computation time. To overcome the drawbacks, we propose a new solution based on the Adaboost algorithm and the Back Propagation Network (BPN) of a Neural Network (NN), combining local and global features with cascade architecture to detect human faces. We use the Modi ed Census Transform (MCT) feature, which belongs to texture features and is less sensitive to illumination, for local feature calculation. In this approach, it is not necessary to preprocess each sub-window of the image. For classi cation, we use the structure of the hierarchical feature to control the number of features. With only MCT, it is easy to misjudge faces and, therefore, in this work, we include the brightness information of global features to eliminate the False Positive (FP) regions. As a result, the proposed approach can have a Detection Rate (DR) of 99%, an FPs of only 11, and detection speed of 27.92 Frames Per Second (FPS).

Traffic counts are one of the fundamental data sources for a variety of transportation applications ranging from assessment of current transportation system conditions to future transportation planning and forecasting. A variety of traffic data collection methods have been used to provide continuous traffic count coverage at selected locations to estimate annual average daily traffic. This paper evaluated the performance of Video Detection System.Video Detection System was investigated under various conditions including mounting styles, heights, and roadway offsets. This paper conducted a lane-by-lane analysis and the results indicated that Video Detection System data present reasonably accurate data, although these data exhibit more variability compared to Automatic Traffic Recorder data. The paper provides an analysis of the strengths and weaknesses of the Video Detection System technology based data as compared to the Automatic Traffic Recorder data and helps in the decision process of whether to use the data for specific transportation planning and strategic decision applications.

The Least Squares (LS) approach has been widely used for solving GPS navigation problems. Despite its many superior properties, however, the LS estimate can be sensitive to outliers, and its performance, in terms of accuracy and statistical inferences, may be compromised when the errors are large and heterogeneous. The GPS signal is strongly a ected by multipath propagation errors. The LS is not able to cope with the above condition to provide a useful and plausible solution. In this paper, an alternative approach, based on the Least Absolute Deviation (LAD) criterion, for estimating navigation solutions is carried out. As a robust estimator, the LAD estimator is known to approximately produce maximum-likelihood estimation. In this case, the maximum-likelihood estimator is obtained by minimizing the mean absolute deviation, rather than the mean square deviation, and, accordingly, can perform robust and e ective estimations. Unlike the LS method, the LAD method is not as sensitive to outliers and, so, may provide more robust estimates. Therefore, the LAD method provides a useful and plausible navigation solution. Simulation results show that the method can e ectively mitigate GPS multipath errors.

This paper proposes an energy-aware routing method, P-AODV. Based on the AODV routing protocol, this study designed a Passive Power Control (PPC) algorithm to enhance energy eciency. The proposed method includes signal detection and power setting phases. During the signal detection phase, the source node broadcasts the route request packets (RREQ), and the downstream intermediate node calculates the optimum power level at the upstream intermediate node according to the Received Signal Strength Indicator (RSSI) after receiving the RREQ. During the power set phase, the destination node returns the RREQ packet to notify the power level of the upstream nodes to the source node. After that, all nodes from the source node to the destination node, engaged in transmission, will transmit data at a coordinated power level. As indicated by the computer simulation results, P-AODV could have 30% higher energy eciency than traditional AODV. This PPC algorithm could also be used in Wireless Mesh Networks (WMNs).

Transportation agencies have recently been struggling to provide rapid roadway construction and repair work with minimal disruption to traffic and maintenance over the ser-vice lifetimes. Precast members provide the characteristics of a controlled roadway quality and rapid construction while the filler between the slabs can be damaged due to the penetra-tion of water and debris. Also, vertical anchor bolts connecting the slabs and crossbeams are associated with unexpected displacement. In this study, a finite element analysis is conducted to develop a modular roadway slab. Two preliminary analysis models are proposed with vary-ing boundary conditions. A combination of various loads was used to determine the appropri-ate boundary conditions. The model with pinned supports at both ends was selected after comparing the displacements and stresses. Since the model had high horizontal force due to the boundary conditions, pinned at both end of the slab, a new system was developed to re-duce the horizontal force by allowing small horizontal displacements at the supports while still fastening vertically. The results of a detailed model analysis showed that the displacement al-lowance with a horizontal filler material should be about 5 mm and the anchor bolts require horizontal shear force of approximately 80 kN.

This study proposes a vehicle lane departure warning system (LDWS) that sends warnings to the driver by using the road marking recognition system of two cameras that are installed on the left and right sides of a vehicle. LWDS aims to solve various problems that may possibly arise when driving on the road. LWDS mainly warns the driver about the unexpected conditions. This system generates warnings by using cameras to capture images continuously by identifying the movement direction of a vehicle using the left and right road markings and predicting the driving direction of the vehicle. Two cameras are installed on the right and left side mirrors of the vehicle to promote Advanced Driver Assistance Systems (ADAS). The left mirror-view camera detects the movement on the right lane, whereas the right mirror-view camera detects the movement on the other lane. These two cameras detect the movement of a vehicle using algorithms and send the driver warning signals independently. Our algorithm combines these signals to analyze environment situations. The used algorithms include brightness adjustment, binarization, dilation, erosion, and edge detection image processing techniques. We tested the LUXGEN M7 in an experiment to verify the feasibility of our proposed system.

This paper presents an anti-shake system to improve the quality of digital dynamic image sequences. The achievement of this study is to develop a digital dynamic image stabilization system using basic image processes and a signal-smoothing algorithm. The dynamic image stabilization method includes three steps: motion vector detection, motion vector smoothing, and image sequence reconstruction. The rst step is to estimate the motion vector of dynamic image sequences using the three-step method. The second step is to separate involuntary motion vectors from motion vectors by the Fourier linear combiner algorithm. The last step is to compensate the unstable dynamic image sequences according to the involuntary motion vector and to build stable dynamic image sequence. The performance of the smoothing algorithm is also discussed in this paper. The Smooth Index (SI) is used to evaluate the performance of the anti-shake system presented in this paper. From the results of the experiments under some di erent conditions, the SI of the digital dynamic image sequences can reach 2.224 through the anti-shake system presented in this paper.

To evaluate the shear performance of deep beams reinforced with Glass Fiber Reinforced Polymer (GFRP) plate, a test was conducted on 8 specimens. Test variables included reinforcement, shear span ratio, area of reinforcement, and e ective depth. The e ects of the test parameters on the shear strength of the test specimens were evaluated. The test result showed that smaller span ratio leads to larger shear strength, and that increase in the area of reinforcement and e ective length increased shear strength. All test results were compared with strut-and-tie models suggested by ACI 318 and CSA.

This paper propose an efficient VLSI architecture of camera distortion correction based on neural camera distortion model (NCDM). Different from conventional imaging method uses over two kind models to correct the camera and lenses distortions, the NCDM uses a single model to correct at once the geometry distor-tion and the unsymmetrical manufacture errors. The NCDM with four neurons perform the wide-angle dis-tortion correction, results show that the maxima whole l corrected error in a image is less than 1.1705 pixels, and that the MSE approaches 0.1743 between the corrected and ideal results. The distortion correction by NCDM is 429x more accurate than the conventional approach. The chip size of NCDM is 1.51x1.51 mm2 that contains 126K gates by using TSMC 90 nm CMOS technology process. As working at 240Mhz, this ar-chitecture can correct 30 frames and Full-HD resolution video per second. Results show that the maximal corrected error in a whole image is less than 1.4 pixels, and that the mean square error approaches 0.0376 be-tween the corrected and ideal results.

In brushless Interior Permanent-Magnet (IPM) motors there exists an oscillatory torque that is induced by the mutual interaction of permanent magnets mounted on the rotor and a slotted structure formed on the stator, generally called the cogging torque. This undesirable torque mainly causes vibration, position inaccuracy and acoustic noise from brushless IPM motors. This paper investigates the in uence of geometric parameters on the cogging torque of brushless IPM motors. An exterior-rotor brushless IPM motor, with embedded magnet poles in a V-shape, is introduced. With the aid of commercial nite-element analysis software, cogging torque waveforms of brushless IPM motors are accurately calculated. The e ects of geometric parameters on the cogging torque, including magnet span angle of the rotor, shoe depth and shoe ramp of the stator, dummy slots notched on the stator, depth of the dummy slot and dummy slots notched on both the stator and rotor, are discussed. Ten design cases, with di erent values of design parameter, are presented to e ectively mitigate the cogging torque. Design case X of the brushless IPM motor, with dummy slots on the stator, performs better than the original brushless IPM motor, with a 79.1% decrease in the peak value of the cogging torque.

In the paper, a cloud-dust based intelligent maximum performance analysis system for power generation with solar energy is proposed. In order to resolve performance problems for power generation using solar energy, factors of the photovoltaic are integrated to the cloud-dust based intelligent maximum power analysis system for computing. This study concerns the development of a maximum performance analysis system for power generation using solar energy, in order to improve the e ects of di erent regions on the solar panels and enable them to get maximum eciency of power generation. The design methodology of this study includes: (1) Records and surveillance module; (2) Prediction and assessment module; (3) Performance diagnosis module; and (4) Maintenance prescription module, with which we are able to nd the design and implementation of records, surveillance, prediction, assessment, diagnosis and prescription for power generation with solar energy. It has worked successfully. The advantages of the clouddust based intelligent, maximum performance analysis system for power generation with solar energy include an increase in the overall competitive performance of the products, and a reduction in the cost of the products and use of human resources.

This paper proposes a storage-independent model for SPARQL-to-SQL translation algorithms based on a relational view. In the development of Web ontology research, the translation from SPARQL to SQL continues to be an issue. Previous research has focused on an ecient and complete translation from SPARQL queries to equivalent SQL queries. However, these translation algorithms depend on speci c storage structures. When we modify the storage structure, the translation algorithm should also be modi ed to suit the changed storage structure. This has motivated study of the issue of a model for the independent use of storage structures by algorithms. These can then guarantee independence between translation algorithms and storages by generating relational views, and improve the application and usability of the translation algorithm. In addition, this paper presents experiment results showing the accuracy and no data loss rate of query results for di erent storages.

This paper presents a novel computational model of stereovision for improving the accuracy of three-dimensional data extracted from a stereo-pair image with no e ect of changes in focal length. For decades, most previous studies on stereovision have focused on the establishment of stereo matching, and have made conclusions on the premise of a xed focus. In general, error in the depth estimate becomes bigger when the focus and aperture are unknown or not xed. For that reason, a three-stage framework is proposed in this paper to modify the conventional stereovision model for improving the accuracy of depth estimation. The rst stage is to modify the computational model of conventional stereovision for varifocal cameras. Then, the spacing of depth intervals in the non-uniform spacing of discrete depth levels can be altered, in particular, to be una ected by changes in focal length. Finally, by considering the ane transformation, we add the deformation coecient into the modi ed stereovision model for correcting three-dimensional ane deformations. Experimental results demonstrated that the depth estimation from stereo images using the proposed scheme was more accurate than conventional methods. The percentage error of most estimates fell between 0.06%-0.82%, and the error value increased from 0.02 cm to 2.21 cm within 6 m.

In a Social Network Service (SNS), a large amount of data with a variety of characteristics is generated through voluntary participation of users. These data are called Big Social Data. «Big social data can identify not only content registered on the web but also the relations of the friends of users. One of the most representative studies on SNS is analysis of the characteristics of social content and social relations, because SNS users tend to add people who are in close contact with them and have similar interests to their list of friends. Finding new knowledge from these large amounts of big social data can be very useful. This paper proposes a polarity analysis method for discovering hidden knowledge based on formal concept analysis in SNSs called PA-DHK. Further, we show, via experiments, that our data analysis approach can be applied to knowledge discovery using association rules.

HEVC is a new video coding standard, which has been developed by the JCT-VC (Joint Collaborative Team on Video Coding) for applications with ultra high de nition video coding. In HEVC, there are three types of encoding block: the Coding Unit (CU), the Predict Unit (PU) and the Transform Unit (TU). If every block of video pixels goes through the complete mode decision process, a large amount of encoding time is required, limiting its applicability in real-time applications. This paper proposes a fast algorithm that can reduce signi cantly the computation time of intra coding for each Coding Unit (CU) block. In the proposed algorithm, depth information of the Coding Block Tree (CBT) from the previously encoded frame is used to help determine the current CU split depth, which can skip many unnecessary mode computations for some cases of block decomposition. Experimental results by HM10.1 show that the proposed method can provide about 20% time-saving and maintain a negligible degradation of coding eciency.

The objective of this research is to use modi ed digital halftoning techniques of Amplitude Modulation (AM) and Frequency Modulation (FM) to integrate lenticular lenses with switching images and embedded infrared watermarks. Modi ed digital halftoning techniques have been used to compose animated graphics and the infrared watermark. The infrared watermark consists of varied combinations of AM/FM or FM/FM halftone dots in black/cyan, magenta and yellow inks. Since the carbon material in black ink can absorb infrared light, the hidden watermark becomes visible under infrared detection. Using a lenticular lens, the image-switching e ect can be achieved by viewing from di erent orientations. The results show that FM/FM halftone dots have a better performance in implementing a lenticular display with both image-switching features and hidden watermarks that can be observed under infrared light. The technique developed in this research can provide value-added applications for anti-forgery and product protection.

The settlement process and the biological treatment process of raw water puri cation can often not be completed eciently under low ambient temperature in cold rural areas. Thus, a supplementary heating system is essential for water puri cation structures under these conditions. This paper proposes a novel hybrid heating system with a combination of solar heating and straw-boiler heating, in which the solar collectors and the straw boiler run in parallel as the heating source. The heat transfer model of the coagulation-sedimentation tank, the main heating position in a puri cation structure, has been established. Employing FLUENT software, the heat transfer conditions of di erent heating modes are simulated and analyzed to determine the optimal heating location. Then, the e ect of the insulation thickness on heat loss is simulated, and, thus, the optimal thickness of the insulation layer is determined. Finally, for the novel heating system, the solar fraction and the duty ratio of the straw boiler in di erent months are calculated. Results show that the proposed heating system can perfectly meet the temperature demands of a water treatment plant in cold rural areas, and provide theoretical guidance and technical support for water puri cation structure heating in cold rural areas.