Applied Mechanics and Materials Vols. 284-287

Abstract: With the constantly increasing public availability of high resolution satellite imagery, interest in automatic road extraction from this imagery has recently increased. Road extraction from high resolution satellite imagery refers to reliable road surface extraction instead of road line extraction because roads in the imagery mostly correspond to an elongated region with a locally constant spectral signature rather than traditional thin lines. This paper proposes a novel automatic road extraction approach that is based on a combination of image segmentation and one-class classification and consists of two main steps. First, the image is segmented using a modified previous segmentation algorithm to achieve more reliable segmentation for road extraction. The key road objects are then automatically extracted from the segmented image to obtain road training samples. Then one-class classification, based on a support vector data description classifier, is carried out to extract the road surface area from the image. The experimental results from a pan-sharpened KOMPSAT-2 satellite image demonstrate the correctness and efficiency of the proposed method for its application to road extraction from high resolution satellite image.

Abstract: Vision based band gesture recognition provides a more nature and powerful means for human-computer interaction. A fast detection process of hand gesture and an effective feature extraction process are presented. The proposed a hand gesture recognition algorithm comprises four main steps. First use Cam-shift algorithm to track skin color after closing process. Second, in order to extract feature, we use BEA to extract the boundary of the hand. Third, the benefits of Fourier descriptor are invariance to the starting point of the boundary, deformation, and rotation, and therefore transform the starting point of the boundary by Fourier transformation. Finally, outline feature for the nonlinear non-separable type of data was classified by using SVM. Experimental results showed the accuracy is 93.4% in average and demonstrated the feasibility of proposed system.

Abstract: In recent years, the grey system theory is widely applied in academic community, and it has obtained very good results. This study proposes GSP chart (Grey Student-Problem chart) which combines S-P chart (Student-Problem chart) with GRA (Grey Relational Analysis). This paper selects the students whose Gamma value is near 0.5 from the GSP chart. Those students are defined as the Misconception-Students. Then we analyze those students’ answers and select the right and wrong answers created by those Misconception Students at the different time. It is said that the misconception questions are defined as the Misconception-Problem Chart. Through the Problem-Concept Chart conducted by teachers, we can define the Misconception Rate and the Misconception Order by comparing the Misconception-Concept Chart. This method can not only accurately predict where students’ misconception is, but also apply to the remedial instruction.

Abstract: Color space conversion has become a very important role in image and video processing systems. To speed up some processing processes, many communication and multimedia video compression schemes use luminance-chrominance type color spaces, such as YCbCr or YUV, making a mechanism for converting between different formats necessary. Therefore, techniques which efficiently implement this conversion are desired. For the recent years, a new field of research called Evolvable Hardware (EHW) has emerged which combines aspects of evolutionary computation with hardware design and synthesis. It is a new scheme inspired by natural evolution, for automatic design of hardware systems. This paper presents a novel evolutionary approach for efficient implementation of a RGB to YCbCr color space converter suitable for Field Programmable Gate Array (FPGAs) and VLSI. In the proposed method, we use the genetic algorithm to automatically evolve the multiplierless architecture of the color space converter. The architecture employs only a few shift and addition operations to replace the complex multiplications. The experimental results represent that the performance of implemented architecture is good at RGB to YCbCr color space converting, and it also has the advantages of high-speed, low-complexity, and low-area.

Abstract: This study adopts popular back-propagation neural network to make one-period-ahead prediction of the stock price. A model based on Taylor series by using both fundamental and technical indicators EPS and MACD as input data is built for an empirical study. Leading Taiwanese companies in non-hi-tech industry such as Formosa Plastics, Yieh Phui Steel, Evergreen Marine, and Chang Hwa Bank are picked as targets to analyze their reasonable prices and moving trends. The performance of this model shows remarkable return and high accuracy in making long/short strategies.

Abstract: This paper refers to a novel (r, n)-threshold secret image sharing scheme with low information overhead. The secret image is encoded into n noise-like shadow images in such a way that any r of the n shares can be used to reveal the secret, and no information about the secret can be revealed from any r–1 or fewer shares. The size of the shadow images is relatively small. Compared with the commonly used in the field of secret image sharing “Thien-Lin algorithm (2002),” the proposed scheme provides an alternative solution for light images. For the security analysis in the case of a 256x256 gray level secret image, if a hacker acquires any r – 1 shadow images, the hacker can construct only r – 1 equations, then the possibility of guessing the right solution is only 1/256. Hence, there are (256x256)/r polynomials, the possibility of obtaining the right image is only (1/256) (256x256)/r. The experimental results and theoretically analysis demonstrate that the proposed scheme performs well.

Abstract: The Eigen-FLS using an eigenspace-based scheme to build up fuzzy logic system (FLS) fast for speech pattern recognition applications has been developed in the author’s previous works. However, speech pattern recognition by Eigen-FLS will still encounter a dissatisfactory recognition performance when the collected data for eigen value calculations of the FLS eigenspace, i.e. the eigen-decomposition process, is scarce. To regulate the influence of Eigen-FLS when data from a test speaker for eigen-decomposition is improper, this paper proposes an EigenMLLR-like Eigen-FLS approach. The developed EigenMLLR-like Eigen-FLS integrates the kernel idea of EigenMLLR speaker adaptation for properly adjusting the target speaker’s Eigen-FLS model in the eigenspace of FLS. EigenMLLR-like Eigen-FLS developed in this paper will be more robust than conventional Eigen-FLS in a speech pattern recognition application with an adverse condition of insufficient data from the speaker.

Abstract: At present, the synthesizing faces of different ages does not emphasize on feature alignment and rectification of twisted images. If these situations do happen, they might cause failure and inaccuracy on synthesizing images. In this paper, we propose a reversible human facial aging/rejuvenating synthesis system which is implemented by Active Shape Model (ASM) integrated with Log-Gabor Wavelet, which can be used to search for the dementia elderly. First, we use AdaBoost and ASM algorithm to extract the feature set of human face, and rectify them by the concept of facial geometric invariance. The invariant concepts are the distance between inner corners of both eyes and the distance between the nose and chin. Then, we find manually one target image which is similar to the test image from the database, and analyze age texture of this human image by Log-Gabor wavelet in order to retrieve decomposition maps. Finally, we can effectively simulate human facial images of people of different ages by controlling the number of decomposition map of images and objectively judge the results via the density of wrinkles.

Abstract: It is generally believed that the same piece of music, after a change of melody or associated instrument, could often elicit a different emotion, thus resulting in a change of music style. However, so far there have been rarely systematic studies on the relationships among the tunes, instruments, and aroused feelings. This paper proposes a framework to perform an automatic style conversion for the music represented in a MIDI format. Compared with most existing tools for music compositions where a professional background is usually assumed, we aim to provide a user-friendly system such that, with simply a mouse click, a desired music style transformation could be achieved by automatically adjusting the associated tempo and instruments of the music. Numerous music styles have been tested and results are quite satisfactory. We believe such a framework could be standardized and adopted in nowadays portable devices, such as laptops, tablet PCs, PDA, or smart phones.

Abstract: For Chinese information processing, automatic classification based on a large-scale database for different patterns of semantic word-formation can remarkably improve the identification for the unregistered word, automatic lexicography, semantic analysis, and other applications. However, owing to noise, anomalies, nonlinear characteristics, class-imbalance, and other uncertainties in word-formation data, the predictive performance of multi-criteria optimization classifier (MCOC) and other traditional data mining approaches will rapidly degenerate. In this paper we put forward an novel MCOC with fuzzification, kernel, and penalty factors (FKP-MCOC) based on layered and weighted graph edit distance (GED): firstly the layered and weighted GEDs between each semantic word-formation graph and prototype graphs are calculated and used for the dissimilarity measure, then the normalized GEDs are embedded into a new feature vector space, and FKP-MCO classifier based on the feature vector space is built for predicting the patterns of semantic word-formation. Our experimental results of Chinese word-formation analysis and comparison with support vector machine (SVM) show that our proposed approach can increase the separation of different patterns, the predictive performance of semantic pattern of a new compound word.