K-Means Clustering Optimizing Deep Stacked Sparse Autoencoder
AbstractBecause of the large structure and long training time, the development cycle of the common depth model is prolonged. How to speed up training is a problem deserving of study. In order to accelerate training, K-means clustering optimizing deep stacked sparse autoencoder (K-means sparse SAE) is presented in this paper. First, the input features are divided into K small subsets by K-means clustering, then each subset is input into corresponding autoencoder model for training, which only has fewer nodes in the hidden layer than traditional models. After training, each autoencoder ’s trained weights and biases is merg...
Source: Sensing and Imaging - February 10, 2019 Category: Biomedical Engineering Source Type: research
Pointwise Multi-resolution Feature Descriptor for Spectral Segmentation
AbstractIn the process of spectral segmentation, it is crucial to compute a reliable affinity matrix with different features of an image. In this paper, we present a method of constructing the affinity matrix based on multi-resolution features extracted from the original features. A pointwise multi-resolution feature descriptor (PMFD) is designed based on spectral graph wavelets, which characterize the topology of the image centered at different pixels. After choosing the scales of interest in our descriptor, a new affinity matrix is constructed based on the extracted features. For large-size affinity matrixes, it is diffi...
Source: Sensing and Imaging - January 25, 2019 Category: Biomedical Engineering Source Type: research
Multiple Proposals for Continuous Arabic Sign Language Recognition
AbstractThe deaf community relies on sign language as the primary means of communication.
For the millions of people around the world who suffer from hearing loss, interaction with hearing people is quite difficult. The main objective of sign language recognition (SLR) is the development of automatic SLR systems to facilitate communication with the deaf community. Arabic SLR (ArSLR) specifically did not receive much attention until recent years. This work presents a comprehensive comparison between two different recognition techniques for continuous ArSLR, namely a Modified k-Nearest Neighbor which is suitable for sequ...
Source: Sensing and Imaging - January 17, 2019 Category: Biomedical Engineering Source Type: research
Recognition of Weld Penetration During K-TIG Welding Based on Acoustic and Visual Sensing
AbstractIn the field of welding process control, on-line monitoring of welding quality based on multi-sensor information fusion has attracted more attention. In order to recognize the penetration state of the Keyhole mode Tungsten Inert Gas welded joint in real time, an acoustic and visual sensing system was established in this paper. The acoustic and visual features that characterize the penetration state of the welded joints in 34 dimensions were extracted and the variation of the acoustic signal and the keyhole geometry were analyzed. In addition, the weighted scoring criterion based on the Fisher distance and the maxim...
Source: Sensing and Imaging - January 2, 2019 Category: Biomedical Engineering Source Type: research
Decluttering Using Wavelet Based Higher Order Statistics and Target Detection of GPR Images
AbstractImages from ground penetrating radar (GPR) may be obscured by high clutter noise over the target signal, making target detection difficult. In this contribution, a decluttering technique is applied to GPR images. This clutter removal/reduction is achieved through wavelet decomposition and application of two methods using the third and fourth order statistics: skewness and kurtosis. These higher order statistics remove clutter but retain target signatures. Different scenarios are considered for real GPR images collected in our controlled lab environment set up and peak signal to noise ratio are compared for the two ...
Source: Sensing and Imaging - December 10, 2018 Category: Biomedical Engineering Source Type: research
Modeling and Enhancement of Piezoelectric Accelerometer Relative Sensitivity
AbstractThe piezoelectric accelerometer is an electronic instrument based on the direct effect of the piezoelectric material, this device is widely used in the industries to monitor and detect defects of rotating machines in an early stage. In this paper, a thorough study of the piezoelectric accelerometer is carried out to understand its design and operation principle. A mathematical model of the accelerometer is developed based on Newton motion law then a new relative sensitivity equation in function of measurement error is extracted. This new equation has allowed a significant reduction in the measurement error, a maxim...
Source: Sensing and Imaging - December 4, 2018 Category: Biomedical Engineering Source Type: research
On Stability Analysis of a Class of Nonlinear Systems with a Focus on Composite Nonlinear Feedback Approach
AbstractObtaining a desired transient performance of closed-loop system is the most important issues in impractical and industrial applications, in general. It is to note that the small overshoot and acceptable settling time of the system response are almost two typical characteristics in the transient performance. There are a number of contradictions between these performances; and therefore, a tradeoff between these ones should deeply be considered. In a word, a so-called composite nonlinear feedback (CNF) method is an efficient and simple technique which is employed to overcome the contradiction of simultaneous achievem...
Source: Sensing and Imaging - October 10, 2018 Category: Biomedical Engineering Source Type: research
Plasmonic Biosensor in NIR with Chalcogenide Glass Material: On the Role of Probe Geometry, Wavelength, and 2D Material
AbstractSamarium doped chalcogenide core, perfluorinated polymer clad with Ag metal and plasmonic 2D materials based plasmonic fiber-optic sensor is simulated and analyzed in near-infrared (NIR) wavelength regime. Proposed sensor is directed at the detection of the malignancy stages of liver tissues. The performance analysis (in terms of figure-of-merit,i.e., FOM) was carried out taking into account the MoS2 and graphene monolayers as performance enhancing 2D materials. The analysis suggests that FOM values of MoS2-based sensor probe are better than graphene-based probe. Further, a comparative study shows that fiber-optic ...
Source: Sensing and Imaging - October 3, 2018 Category: Biomedical Engineering Source Type: research
Color Image Modification with and without Hue Preservation
AbstractColor image modification is an essential component for several applications and the grayscale transformation is generally mapped to the color image indirectly. Although several techniques have been used for this transfer, they suffer from gamut mapping issue. In this paper, it is aimed to study the mapping of grayscale transformations to the color scale in different perspectives. Modifying the image in different color space than the original retains hue to a promising extent, but suffers from the gamut problem. A generic scheme to map grayscale changes to the color space for all kinds of spatial modification is pro...
Source: Sensing and Imaging - August 17, 2018 Category: Biomedical Engineering Source Type: research
Recognizing Anomalies in Urban Road Scenes Through Analysing Single Images Captured by Cameras on Vehicles
AbstractIn this paper, we propose to recognize anomalies in urban road scenes through analysing single images captured by cameras on driving vehicles. Anomaly detection is one of the most important functions for visual driver-assistance systems and autonomous vehicles. Anomaly detection provides drivers and autonomous vehicles important information about driving environments and help them drive more safely. In this work, we define anything on roads that are within a certain distance of a driving vehicle and pose potential dangers for it as anomalies, such as traffic accidents, reckless driven vehicles and pedestrians. The ...
Source: Sensing and Imaging - August 9, 2018 Category: Biomedical Engineering Source Type: research
A Just Noticeable Difference-Based Video Quality Assessment Method with Low Computational Complexity
AbstractA Just Noticeable Difference (JND)-based video quality assessment (VQA) method is proposed. This method, termed as JVQ, applies JND concept to structural similarity (SSIM) index to measure the spatial quality. JVQ incorporates three features, i.e. luminance adaptation, contrast masking, and texture masking. In JVQ, the concept of JND is refined and more features are considered. For the spatial part, minor distortions in the distorted frames are ignored and considered imperceptible. For the temporal part, SSIM index is simplified and used to measure the temporal video quality. Then, a similar JND concept which compr...
Source: Sensing and Imaging - August 6, 2018 Category: Biomedical Engineering Source Type: research
Lock-in Amplifier Based Eddy Current Instrument for Detection of Sub-surface Defect in Stainless Steel Plates
AbstractFor detection of sub-surface defects using eddy current (EC) method, increasing the depth of penetration of ECs is essential. This can be achieved through strengthening of the primary magnetic field from the EC probe. This can be accomplished by using low-frequency high amplitude excitation current, precise phase lag measurement and high throughput probes. Working on these lines, the paper presents development of lock-in amplifier based EC instrument and cup-core send-receive type probe. Experiment results indicate that the proposed instrument and probe is able to detect sub-surface defects located at 8.0 mm belo...
Source: Sensing and Imaging - August 6, 2018 Category: Biomedical Engineering Source Type: research
