Catálogo de publicaciones - libros

Research on the merge area has mainly dealt with free flow traffic and research on the congested traffic at the merge area is rare. This study investigates the relationship between mainline traffic and on-ramp traffic at three different segments of the merge area. For this purpose, new indicators based on traffic variables such as flow, speed, and density are used. The results show that a negative relationship exists between mainline and on-ramp flow. It is also found that the speed and the density of the right two lanes in the mainline traffic are significantly affected by the on-ramp flow. Based on the correlation analysis of the indicators, it is confirmed that the right two lanes of the freeway mainline are influenced by the ramp flow. The revealed relationships between mainline and on-ramp traffic may help to analyze the capacity of the downstream freeway segment of the merge area in congested traffic.

A novel family of stochastic methods developed for function recovery tasks is presented and its properties are discussed in some detail. A new image resize facility based on these new methods is applied to an image and this compares favorably in quality to the application to this image of an equivalent facility from a popular commercial graphics package.

An automatic face analysis system is proposed which uses face recognition and facial physiognomy. It first detects human’s face, extracts its features, and classifies the shape of facial features. It will analyze the person’s facial physiognomy and then automatically make an avatar drawing using the facial features. The face analysis method of the proposed algorithm can recognize face at real-time and analyze facial physiognomy which is composed of inherent physiological characteristics of humans, orientalism, and fortunes with regard to human’s life. The proposed algorithm can draw the person’s avatar automatically based on face recognition. We conform that the proposed algorithm could contribute to the scientific and quantitative on-line face analysis fields as well as the biometrics.

Moving cast shadows on object distort figures which causes serious detection deficiency and analysis problems in ITS related applications. Thus, shadow removal plays an important role for robust object extraction from surveillance videos. In this paper, we propose an algorithm to eliminate moving cast shadow that uses features of color information about foreground and background figures. The significant information among the features of shadow, background and object is extracted by PCA transformation and tilting coordinates system. By appropriate analyses of the information, we found distributive characteristics of colors from the tilted PCA space. With this new color space, we can detect moving cast shadow and remove them effectively.

Molecular imaging can detect abnormal functions of living tissue. Functional abnormality in gene expression or metabolism can be represented as altered volume or probe intensity. Accurate measurement of volume and probe intensity in tissue mainly relies on image segmentation techniques. Thus, segmentation is a critical technique in quantitative analysis. We developed an automatic object marker-driven three dimensional(3D) watershed transform for quantitative analysis of functional images. To reduce the discretization error in volume measurement less than 5%, the size criteria for digital spheres were investigated to provide the minimum volume. When applied to SPECT images, our segmentation technique produced 89% or higher accuracy in the volume and intensity of tumors and also showed high correlation with the ground truth segmentation (> 0.93). The developed 3D method did not require interactive object marking and offered higher accuracy than a 2D watershed approach. Furthermore, it computed faster than the segmentation technique based on the marker-driven gradient modification.

In this paper, the transmission service for medical image is proposed via IEEE 802.15.4a on WPAN environment. Also, transmission and receiving performance of medical image using TH UWB-IR system is evaluated on indoor multi-path fading environment. On the results, the proposed scheme can solve the problem of interference from the medical equipment in same frequency band, and minimize the loss due to the indoor multi-path fading environment. Therefore, the transmission with low power usage is possible.

Image based spam email can easily circumvent widely used text based spam email filters. More and more spammers are adapting the technology. Being able to detect the nature of email from its image content is urgently needed. We propose to use OCR (optical character recognition) technology to extract the embedded text from the images and then assess the nature of the email by the extracted text using the same text based engine. This approach avoids maintaining an extra image based detection engine and also takes the benefit of the strong and reasonably mature text based engine. The success of this approach relies on the accuracy of the OCR. However, regardless of how good an OCR is, misrecognition is unavoidable. Therefore, a Markov model which has the ability to tolerate misspells is also proposed. The solution proposed in this paper can be integrated smoothly into existing spam email filters.

There are several sub-optimal search techniques for fast algebraic codebook search of ACELP speech codecs. Focused search method, depth-first tree search method and pulse replacement methods are used to reduce computational complexity of algebraic codebook search. In previous pulse replacement methods, the computational load is increased as the pulse replacement procedure is repeated. In this paper, we propose a fast algebraic codebook search method based on iteration-free pulse replacement. The proposed method is composed of two stages. At the first stage, an initial codevector is determined by the backward filtered target vector or the pulse-position likelihood-estimate vector. At the second stage, after computing pulse contributions for every track the pulse replacement is performed to maximize the search criterion over all combination replacing the pulses of the initial codevector with the most important pulses for every track. The performance of the proposed algebraic codebook search method is measured in terms of the segmental signal to noise ratio (SNRseg) and PESQ (Perceptual Evaluation of Speech Quality) using various speech data. Experimental results show that the proposed method is very efficient in computational complexity and speech quality comparing to previous pulse replacement methods.

This paper presents a combined handsfree speech enhancement method based on a spatialpost-filter. The scheme uses a linear microphone array to capture a speech signal that has been corrupted by babble noise, car noise, and interference signals. Simulation results for real environment show that the proposed structure achieves a maximum interference suppression of 12 dB, an improvement of 6 dB over the delay and sum beamformer. Furthermore, the system is robust in the presence of distortion as opposed to the generalized sidelobe canceller. The subjective evaluation has shown that the combined system of delay and sum with the minimum mean square error estimator using a noncausal signal to noise ratio (SNR) estimator obtained 3.8 points on a fivepoint.

In this paper, we propose a bandwidth extension (BWE) algorithm of a narrowband speech coder for music delivery services over IP networks. The proposed BWE algorithm is based on an embedded structure of using a baseline coder followed by an enhancement layer. To minimize the bit-rate increase by the enhancement layer, the proposed algorithm shares spectral envelope and excitation parameters between the baseline coder and the enhancement layer. In this paper, we choose the iLBC as the baseline coder and mel-frequency cepstral coefficients (MFCCs) are used to reconstruct higher frequency components at the enhancement layer. By doing this, the bit-rate of the proposed BWE coder is 15.45 kbit/s which is just 0.25 kbit/s higher than the iLBC. We compare the quality of the proposed BWE coder with that of the iLBC, and it is shown from an informal listening test that the proposed BWE coder provides significantly better quality than the iLBC for all four different kinds of music genres such as pop, classical, jazz and rock.