Ibrahim Onaran

Human face detection in video using edge projections

In this paper, a human face detection algorithm in images and video is presented. After determining possible face candidate regions using colour information, each region is filtered by a high-pass filter of a wavelet transform. In this way, edges of the region are highlighted, and a caricature-like representation of candidate regions is obtained. Horizontal, vertical, filter-like and circular projections of the region are used as feature signals in support vector machine (SVM) based classifiers. It turns out that our feature extraction method provides good detection rates with SVM based classifiers.

Falling person detection using multisensor signal processing

Falls are one of the most important problems for frail and elderly people living independently. Early detection of falls is vital to provide a safe and active lifestyle for elderly. Sound, passive infrared (PIR) and vibration sensors can be placed in a supportive home environment to provide information about daily activities of an elderly person. In this paper, signals produced by sound, PIR and vibration sensors are simultaneously analyzed to detect falls. Hidden Markov Models are trained for regular and unusual activities of an elderly person and a pet for each sensor signal. Decisions of HMMs are fused together to reach a final decision.

Pulse Doppler Radar Target Recognition using a Two-Stage SVM Procedure

It is possible to detect and classify moving and stationary targets using ground surveillance pulse-Doppler radars (PDRs). A two-stage support vector machine (SVM) based target classification scheme is described here. The first stage tries to estimate the most descriptive temporal segment of the radar echo signal and the target signal is classified using the selected temporal segment in the second stage. Mel-frequency cepstral coefficients of radar echo signals are used as feature vectors in both stages. The proposed system is compared with the covariance and Gaussian mixture model (GMM) based classifiers. The effects of the window duration and number of feature parameters over classification performance are also investigated. Experimental results are presented.

Detection of underdeveloped hazelnuts from fully developed nuts by impact acoustics

Shell-to-kernel weight ratio is a vital measurement of quality in hazelnuts as it helps to identify nuts that have underdeveloped kernels. Nuts containing underdeveloped kernels may contain mycotoxin-producing molds, which are linked to cancer and are heavily regulated in international trade. A prototype system was set up to detect underdeveloped hazelnuts by dropping them onto a steel plate and recording the acoustic signal that was generated when a kernel hit the plate. A feature vector comprising line spectral frequencies and time-domain maxima that describes both the time and frequency nature of the impact sound was extracted from each sound signal and used to classify each nut by a support-vector machine. Experimental studies demonstrated accuracies as high as 97% in classifying hazelnuts with underdeveloped kernels.

Projections onto convex sets (POCS) based optimization by lifting

Summary form only given. A new optimization technique based on the projections onto convex space (POCS) framework for solving convex and some non-convex optimization problems are presented. The dimension of the minimization problem is lifted by one and sets corresponding to the cost function are defined. If the cost function is a convex function in RN the corresponding set which is the epigraph of the cost function is also a convex set in RN+1. The iterative optimization approach starts with an arbitrary initial estimate in RN+1 and an orthogonal projection is performed onto one of the sets in a sequential manner at each step of the optimization problem. The method provides globally optimal solutions in total-variation, filtered variation, l1, and entropic cost functions. It is also experimentally observed that cost functions based on lp; p <; 1 may be handled by using the supporting hyperplane concept. The new POCS based method can be used in image deblurring, restoration and compressive sensing problems.

Classification of multichannel ECoG related to individual finger movements with redundant spatial projections

We tackle the problem of classifying multichannel electrocorticogram (ECoG) related to individual finger movements for a brain machine interface (BMI). For this particular aim we applied a recently developed hierarchical spatial projection framework of neural activity for feature extraction from ECoG. The algorithm extends the binary common spatial patterns algorithm to multiclass problem by constructing a redundant set of spatial projections that are tuned for paired and group-wise discrimination of finger movements. The groupings were constructed by merging the data of adjacent fingers and contrasting them to the rest, such as the first two fingers (thumb and index) vs. the others (middle, ring and little). We applied this framework to the BCI competition IV ECoG data recorded from three subjects. We observed that the maximum classification accuracy was obtained from the gamma frequency band (65200Hz). For this particular frequency range the average classification accuracy over three subjects was 86.3%. These results indicate that the redundant spatial projection framework can be used successfully in decoding finger movements from ECoG for BMI.

Localization of subthalamic nucleus borders using macroelectrode local field potential recordings.

Deep brain stimulation of the subthalamic nucleus (STN) is a highly effective treatment for motor symptoms of Parkinsons disease. However, precise intraoperative localization of STN remains a procedural challenge. In the present study, local field potentials (LFPs) were recorded from DBS macroelectrodes during trajectory to STN, in six patients. The frequency-vs-depth map of LFP activity was extracted and further analyzed within different sub-bands, to investigate whether LFP activity can be used for STN border identification. STN borders identified by LFPs were compared to border predictions by the neurosurgeon, based on microelectrode-derived, single-unit recordings (MER-SUA). The results demonstrate difference between MER-SUA and macroelectrode LFP recording with respect to the dorsal STN border of -1.00 ±0.84 mm and -0.42 ±1.07 mm in the beta and gamma frequency bands, respectively. For these sub-bands, RMS of these distances was found to be 1.26 mm and 1.06 mm, respectively. Analysis of other sub-bands did not allow for distinguishing the caudal border of STN. In conclusion, macroelectrode-derived LFP recordings may provide an alternative approach to MER-SUA, for localizing the target STN borders during DBS surgery.

Falling Person Detection Using Multi-sensor Signal Processing

Falls are one of the most important problems for frail and elderly people living independently. Early detection of falls is vital to provide a safe and active lifestyle for elderly. In this paper, signals produced by sound and passive infrared (PIR) sensors are simultaneously analyzed to detect suddenly falling elderly people. A typical room in a supportive home can be equipped with sound and PIR sensors. Hidden Markov models are trained for regular and unusual activities of an elderly person and a pet for each sensor signal. Decisions of HMMs can be fused together to reach a final decision.

A hybrid SVM/HMM based system for the state detection of individual finger movements from multichannel ECoG signals

A hybrid state detection algorithm is presented for the estimation of baseline and movement states which can be used to trigger a free paced neuroprostethic. The hybrid model was constructed by fusing a multiclass Support Vector Machine (SVM) with a Hidden Markov Model (HMM), where the internal hidden state observation probabilities were represented by the discriminative output of the SVM. The proposed method was applied to the multichannel Electrocorticogram (ECoG) recordings of BCI competition IV to identify the baseline and movement states while subjects were executing individual finger movements. The results are compared to regular Gaussian Mixture Model (GMM)-based HMM with the same number of states as SVM-based HMM structure. Our results indicate that the proposed hybrid state estimation method out-performs the standard HMM-based solution in all subjects studied with higher latency. The average latency of the hybrid decoder was approximately 290ms.

IDENTIFICATION OF DAMAGED WHEAT KERNELS AND CRACKED-SHELL HAZELNUTS WITH IMPACT ACOUSTICS TIME-FREQUENCY PATTERNS

A new adaptive time-frequency (t-f) analysis and classification procedure is applied to impact acoustic signals for detecting hazelnuts with cracked shells and three types of damaged wheat kernels. Kernels were dropped onto a steel plate, and the resulting impact acoustic signals were recorded with a PC-based data acquisition system. These signals were segmented with a flexible local discriminant bases (F-LDB) procedure in the time-frequency plane to extract discriminative patterns between damaged and undamaged food kernels. The F-LDB procedure requires no prior knowledge of the relevant time or frequency indices of the impact acoustics signals for classification. The method automatically finds all crucial time-frequency indices from the training data by combining local cosine packet analysis and a frequency axis clustering approach, which supports individual time and frequency band adaptation. Discriminant features are extracted from the adaptively segmented acoustic signal, sorted according to a Fisher class separability criterion, post-processed by principal component analysis, and fed to a linear discriminant classifier. Experimental results establish the superior performance of the proposed approach when compared to prior techniques reported in the literature or used in the field. The new approach separated damaged wheat kernels (IDK, pupal, and scab) from undamaged wheat kernels with 96%, 82%, and 94% accuracy, respectively. It also separated cracked-shell hazelnuts from those with undamaged shells with 97.1% accuracy. The adaptation capability of the algorithm to the time-frequency patterns of signals makes it a universal method for food kernel inspection that can resist the impact acoustic variability between different kernel and damage types.