By Topic

Intelligent Signal Processing and Communications Systems (ISPACS), 2012 International Symposium on

Date 4-7 Nov. 2012

Filter Results

Displaying Results 1 - 25 of 180
  • Author index

    Publication Year: 2012 , Page(s): 1 - 60
    Save to Project icon | Request Permissions | PDF file iconPDF (221 KB)  
    Freely Available from IEEE
  • Content list

    Publication Year: 2012 , Page(s): 1
    Save to Project icon | Request Permissions | PDF file iconPDF (333 KB)  
    Freely Available from IEEE
  • Organizing Committee

    Publication Year: 2012 , Page(s): 1 - 7
    Save to Project icon | Request Permissions | PDF file iconPDF (295 KB)  
    Freely Available from IEEE
  • [Front cover]

    Publication Year: 2012 , Page(s): c1
    Save to Project icon | Request Permissions | PDF file iconPDF (333 KB)  
    Freely Available from IEEE
  • Organizing committee

    Publication Year: 2012 , Page(s): 1 - 3
    Save to Project icon | Request Permissions | PDF file iconPDF (199 KB)  
    Freely Available from IEEE
  • Single-shot person re-identification based on improved Random-Walk pedestrian segmentation

    Publication Year: 2012 , Page(s): 1 - 6
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1417 KB) |  | HTML iconHTML  

    Single-shot person re-identification is to match pedestrian images captured from different cameras at different time under the condition of large illumination variations, different viewpoints, and inadequate information of single-shot case. To deal with these challenges, we propose a four-step single-shot person re-identification algorithm that consists of pedestrian segmentation, human region partitioning, feature extraction and human feature matching. Based on an improved Random Walks algorithm, human foreground is segmented by combining the shape prior information and the color seed constraint into the Random Walk formulation. Then color features of HSV histogram and 1-D RGB signal along with texture features from human body parts are used for the person re-identification. The correct match is then determined by the similarity scores of all features with appropriate weight selection. The experimental results demonstrate the superior performance by using the proposed algorithm compared to the previous representative methods. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Human action classification using histogram-based discriminative embedding

    Publication Year: 2012 , Page(s): 7 - 11
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (576 KB) |  | HTML iconHTML  

    In order to have a rich representation for human action, we propose to combine two complementary features so that a human posture can be characterized in more details. In particular, the distance signal feature and the width feature are combined in an effective way to enhance each other's discriminating capability. The resulting feature vector is quantized into mid-level features using k-means clustering. In the mid-level feature space, we apply the nonparametric embedding method to construct a compact yet discriminative subspace model. We have conducted a series of experiments on the Weizmann dataset to validate the proposed scheme. Compared with the existing approaches, our method can achieve high recognition accuracy while having a reduced computational complexity in classification stage. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Scalable near-duplicate video stream monitoring

    Publication Year: 2012 , Page(s): 12 - 15
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (210 KB) |  | HTML iconHTML  

    In this paper, we present a simple but effective algorithm to address the efficiency problem of video stream near-duplicate monitoring in a large-scale repository. In the algorithm, a similarity upper bound between two sequences is calculated incrementally and served as a rough filter to avoid time-consuming computation for the true sequence similarity. It makes the scan process more efficient and guarantees no true positive is missed. Our experiments demonstrate the proposed algorithm yields accurate retrieval result and efficient execution time. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Innovative 3D augmented reality techniques for spinal surgery applications

    Publication Year: 2012 , Page(s): 16 - 20
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (357 KB) |  | HTML iconHTML  

    This paper aims to be an outlet for speeding up workflow of orthopedics surgery and be helpful for surgeons to simply coordinate transformations between several imaging displays. Based on epipolar geometry and simple feature landmarks, a 3-D superimposed imaging approach is developed via the construction of the camera-projector system. The superimposed approach is to rectify the perspective, X-ray and projected image pair using the perspective projection model. The proposed method not only simplifies the computation between surgical instruments and patient for surgeons, but also reduces the radiation exposure. Experimental results for both the synthetic spinal image on dummy and real patient testing have demonstrated the feasibility of our approach. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Large scale 3D scene reconstruction with improved registration of laser range data

    Publication Year: 2012 , Page(s): 21 - 26
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1117 KB) |  | HTML iconHTML  

    The paper proposes two approaches to improve the Iterative Closest Point (ICP) algorithm in the registration of large scale range data obtained by a Velodyne LIDAR at different locations in an outdoor environment. The first proposed approach discards points that cannot be matched in the datasets during the registration process to prevent errors from these points from affecting the results. The second approach extracts feature points that are representative of the datasets to perform the registration process and similarly preventing mismatching points from affecting the results. Experiments show that both approaches perform better than the original ICP algorithm. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An improved Richardson-Lucy algorithm for single image deblurring using local extrema filtering

    Publication Year: 2012 , Page(s): 27 - 32
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1806 KB) |  | HTML iconHTML  

    With the popularity of digital camera, digital image processing is getting more important. One of the most common problems in digital photographing is motion blur. The research in solving the problem of motion blur efficiently is called motion deblur. When taking a photograph, the shaking of camera is the reason causing blurred image. The blur process can be formulated as the image takes convolution operation with the shaking path, which is also known as point spread function. One of the well-known deconvolution algorithms in solving the convolution problem is Richardson-Lucy algorithm. Although Richardson-Lucy can recover the image from blurred image, there are unexpected ringing artifacts in the deblurred image. To solve ringing is the main purpose in recent researches. In our research, by pre-detecting the region and intensity of ringing in the image, we propose an improved Richardson-Lucy algorithm to deblur image and suppress the ringing. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Real-time face detection using Gentle AdaBoost algorithm and nesting cascade structure

    Publication Year: 2012 , Page(s): 33 - 37
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (209 KB) |  | HTML iconHTML  

    In this paper, a face detector based on Gentle AdaBoost algorithm and nesting cascade structure is proposed. Nesting cascade structure is introduced to avoid that too many weak classifiers in a cascade classifier will slow down the face detection speed of this cascade classifier. Gentle AdaBoost algorithm is used to train node classifiers on a Haar-like feature set to improve the generalization ability of the node classifier. Consequently, the face detection performance of the face detector is improved. Experimental results have proved that the proposed algorithm can significantly reduce the number of weak classifiers, increase the detection speed, and slightly raise the detection accuracy as well. On the CIF (352×288) video sequences, the average detection speed of the proposed face detector can achieve 125fps, which is superior to the state-of-the-art face detectors and completely satisfies the demand of real-time face detection. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A color image enhancement method with a high regard for restoration of skin color

    Publication Year: 2012 , Page(s): 38 - 42
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (576 KB) |  | HTML iconHTML  

    Hue (H), saturation (S) and intensity (I) are the attributes of color. Therefore, a lot of color image enhancement techniques are performed in HSI color space. In this paper, a new color image contrast enhancement method is proposed in HSI color space. Histogram equalization (HE) is a popular technique for contrast enhancement of images. It is the most commonly used method due to its simplicity and comparatively better performance on almost all types of images. However, HE shows unnatural results for images with large background. In order to improve this drawback of HE, HE with variable enhancement degree (HE-VED) is introduced. The enhancement degree of HE-VED is controlled by one parameter. Then, HE-VED is applied to saturation and intensity histograms separately. In this paper, we aimed mainly at natural color images include human faces. In the kind of image, it is necessary to describe skin color accurately. Thus, we adjust a parameter of HE-VED, in order to restore the skin color of human faces. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Multimodal person authentication system using features of utterance

    Publication Year: 2012 , Page(s): 43 - 47
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1610 KB) |  | HTML iconHTML  

    In this paper, we propose a multimodal biometrics authentication method using features of an utterance. The proposed authentication method authenticates persons using image and voice signals. Hence, the proposed method can be realized with only a camera and microphone to extract the lip area and voice without the special equipment used in other personal authentication methods and can easily change the registration data. Moreover, the proposed authentication method can provide a key function to the registered phrase of the utterance. In the proposed method, the edges and texture in the mouth are used as image features, and pitch and spectrum envelope are used as voice features. Authentication is realized by classifiers generated by AdaBoost, classifiers are generated for the voice- and image-processing parts. Moreover, each classifier is weighted according to the corresponding confidence and then the final authentication score is calculated. Hence, the proposed method can provide valid authentication results in various environments. Experimental results demonstrate that multimodal processing in the proposed method is more effective than monomodal (only image or voice) processing. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Interactive evolutionary image processing for face beautification using smaller population size

    Publication Year: 2012 , Page(s): 48 - 53
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (529 KB) |  | HTML iconHTML  

    A method to improve image processing with interactive evolutionary computing (IEC) is proposed in order to get a satisfactory output image with smaller population size. IEC is a powerful method for designing a system on the basis of human subjective criteria. Image processing with IEC is effective to process images considering human subjective criteria and taste, however, it requires a large PC system to show enough number of candidate output images on a large display; this feature makes this method difficult to be realized on a small mobile device. Here, the number of the candidates corresponds to the population size in the genetic algorithm (GA) in IEC. If the number of candidate output images is restricted, IE image processing takes more iteration time to get satisfactory result from small number of candidates. In order to solve this problem, a method to improve IEC is proposed so that the initial population is designed effectively by clustering past optimized parameters, and the generation change is modified by considering the significance of the first choice. This method is applied to human face image beautifying system, and the experimental results show that users can get more satisfactory result with smaller number of iterations. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Recognition of fetal facial expressions for neurobehavioral assessment using Relaxing-Mismatch Global Sequence Time Warp Kernel

    Publication Year: 2012 , Page(s): 54 - 57
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (189 KB) |  | HTML iconHTML  

    Most pregnant women are positive feeling and happy when the ultrasound scanning results the normal detection. However, if the finding gives unexpected anomaly detection, it is the cause of unnecessary anxiety and worry. Fetal neurobehavioral assessment is one of the antenatal assessments. The goal is to identify fetal being that is well or at risk and expected that the risk can be prevented or reduced. Fetal behavioral variables, movements and postures, are an indicator of significantly spontaneous activity of fetal central nervous system. Fetal behavioral studies suggest that the frequency of some facial fetal expressions at risk differs from normal. For instance, hydrocephalic fetus has less a fewer eye movement than normal fetuses. Although the individual ultrasound scan has been successively used, the knowledge, the reliability, and the objective are required with the clinical professional experience. The limitation and the human fatigue are often caused of biasing and failing the significant consideration of available information. Thus, automatic facilitated fetal facial expression assessment becomes necessary. Although, Automatic facial expression recognition is widely used in human facial expressions, the system suffers with the time axis distortion. In medical diagnosis, precise recognition system is required. The Relaxing-Mismatch Global Sequence Time Warp Kernel is proposed for natural fetal facial expression recognition. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • 3D model reconstruction based on multiple view image capture

    Publication Year: 2012 , Page(s): 58 - 63
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (476 KB) |  | HTML iconHTML  

    In this paper, a system is designed for three-dimensional model reconstruction. The proposed system uses a camera to take multiple images from the object and analyze those images using structure from motion (SFM) to obtain the camera parameters and sparse point clouds. Therefore, this system can work without any camera calibration prior to the experiments. In order to reconstruct the target object, the patch-based multi-view stereo is used to capture two target postures and compute the three-dimensional point clouds and color information from the target object. Then use the iterative closest point algorithm to register those two target postures. In the last step, the Poisson surfaces reconstruction technique is used to build the three-dimensional mesh model. Finally, the experimental results show that the study acquires proper practicability and robustness. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Lane departure and front collision warning using a single camera

    Publication Year: 2012 , Page(s): 64 - 69
    Cited by:  Papers (2)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (472 KB) |  | HTML iconHTML  

    Improving the driving safety is one major concern for the design of intelligent vehicles. In this paper, we present a monocular vision based driver assistance system for dangerous traffic warning. The video sequences captured from a single camera mounted behind the windshield are used for lane detection and front vehicle identification. Two basic modules, lane departure warning system (LDWS) and front collision warning system (FCWS), are developed and then integrated on an embedded DSP platform for automotive electronics applications. Error analysis on system installation is carried out to verify the correctness of measurements. Experimental results have demonstrated that the proposed technique is able to achieve 97% of accuracy on dangerous traffic warning while maintaining the real-time processing requirement. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A quantitative line-scan image analysis method for colonoscopy diagnosis using FICE

    Publication Year: 2012 , Page(s): 70 - 73
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (281 KB) |  | HTML iconHTML  

    This paper presents a quantitative image analysis technique for colonoscopy diagnosis to distinguish abnormal lesions, especially for using Fuji Intelligent Chromo Endoscopy (FICE). We take the color histogram technique and gray level co-occurrence matrix (GLCM) into account, and proposed an efficient line-scan method for observing early and small tumor during minute diagnosis. The line-scan method is based on semimanual image horizontal line selection of colonoscopy video, then the color histogram index, contrast, energy, homogeneity and entropy are calculated via GLCM matrix of the selection image area. Thus, the method leads the FICE images measurable and provides a view to accelerate and enhance physician to determine the lesions quickly. The results show that the proposed line-scan method significantly evaluates the diagnosis images for the real patient experiments. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • High speed multi-layer background subtraction

    Publication Year: 2012 , Page(s): 74 - 79
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (307 KB) |  | HTML iconHTML  

    Moving object detection is an important and fundamental step for intelligent video surveillance systems, since it provides a focus of attention for post-processing. In this study, a multi-layer codebook-based background subtraction model is proposed to cope with high resolution video sequences for the detection of moving objects. Combining the multi-layer block-based strategy and the feature extraction of blocks, the proposed method can remove most of the background, including non-stationary (dynamic) background, and significantly increase the processing efficiency. Moreover, pixel-based classification is adopted for refining the results of block-based background subtraction, which can further classify pixels as foreground, shadows, and highlights. The experimental results demonstrate that the proposed multi-layer codebook-based background subtraction method (for standard definition video sequences) can provide a high precision and efficient processing speed for moving objects detection. Moreover, extensive experiments have been conducted to compare with various former schemes, and the superiority of the proposed method shows that it can be a very effective candidate for real-time intelligent video surveillance applications. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Hierarchical pedestrian detection under low resolution scenario

    Publication Year: 2012 , Page(s): 80 - 84
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (317 KB) |  | HTML iconHTML  

    The pedestrian detection is a popular research field in recent years, yet the low-resolution issue is rarely discussed for yielding reasonable response time for drivers. In this study, a hierarchical pedestrian detection system is proposed to cope with this issue. In which, two independent features, orientation and magnitude, are adopted as the descriptors to detect pedestrians. Moreover, to meet the real-time requirement, the proposed Probability-based Pedestrian Mask Pre-Filtering (PPMPF) is adopted to initially filter out lots of non-pedestrian regions while retaining as more true pedestrian as possible. In addition, the concept of integral image is also adopted to simplify the calculations of the adopted features. In experimental results, some popular features such as the Haar-like feature and the edgelet feature are adopted for comparison. The results demonstrate that the proposed system offers better performance as well as high processing efficiency, and thus it can be a very competitive candidate for intelligent surveillance applications. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • An improved obstacle detection using optical flow adjusting based on inverse perspective mapping for the vehicle safety

    Publication Year: 2012 , Page(s): 85 - 89
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (637 KB) |  | HTML iconHTML  

    In this paper, we implement a method to detect obstacles around the vehicle by optical flow computation based on inverse perspective mapping. However, this approach had been proposed and used on the vehicle. But this approach is limited on a trajectory of the vehicle which moves along with a straight line. If the trajectory of the vehicle moves along with an arc line, the optical flow values will inconsistent in the left side and in the right side of this image which causes mistakes in obstacle detection. So we propose a method to improve this problem which uses the trajectory of vehicle to calculate the center of turning circle. And the center of turning circle can be used to adjust the inconsistent with optical flow values. After our improvement, the optical flow values can keep consistence even if the trajectory is along with an arc line. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Efficient image interpolation by associating 2nd order local structure and data-adaptive kernel regression

    Publication Year: 2012 , Page(s): 90 - 95
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (373 KB) |  | HTML iconHTML  

    Image interpolation is still a widely studied issue for rescaling a low-resolution image to a high resolution image. This paper tends to modify the data-dependent steering kernel regression image interpolation in order to reduce the computational cost in point-wise determination of data-dependent or nonlinear filter coefficients. Instead solving a kernel-based weighting least mean squared minimization a novel example-based matching approach is introduced and the problem is turned into the nearest neighbor search problem. Through conducted experiments applied to several images the proposed method is verified to reduce the computational time about 50% compared with the steering kernel regression algorithm while almost maintaining image quality. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Speaker dependent visual word recognition by using sequential mouth shape codes

    Publication Year: 2012 , Page(s): 96 - 101
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (401 KB) |  | HTML iconHTML  

    Visual speech recognition or lip reading is an approach for noise robust speech recognition by adding speaker's visual cues to audio information. Basically visual-only speech recognition is applicable to speaker verification and multimedia interface for supporting speaking impaired person. The sequential mouth-shape code method is an effective approach of lip reading for particularly uttered Japanese words by utilizing two kinds of distinctive mouth shapes, known as first and last mouth shapes, appeared intermittently. One advantage of this method is its low computational burden for the learning and word registration processes. This paper proposes a novel word lip recognition system by detecting and determining initial mouth-shape codes to recognize uttering consonants. The proposed method eventually is able to discriminate different words consisting of the same sequential vowel codes though containing different consonant codes. The conducted experiments demonstrate that the proposed system provides higher recognition rate than the conventional ones. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A model update scheme of color-based particle filter for multi-color object tracking

    Publication Year: 2012 , Page(s): 102 - 107
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (625 KB) |  | HTML iconHTML  

    Color-invariance property of objects is effectively employed for developing non-rigid object tracking algorithms in the field of computer vision. This paper develops a novel color-based tracking algorithm for non-rigid 3-D objects with multiple colors. Especially, the proposed particle filter method can track the targets even if the appearance /disappearance of color regions were occurred by self-occlusion and pose variation like self-rotation. The basic idea is closely related to the rules of proximity and common fate in grouping mechanism in human psychology, and a new model updating scheme is introduced. The effective results by the proposed method are obtained by comparing it with other color-based particle filters. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.