Yalcin Albayrak

Object tracking by PI control and image processing on embedded systems

In this study, object tracking application by using image processing and PI controlling is implemented. The color property of the desired object which was wanted to be detected and tracked by image processing technique was utilized and its position in front of the camera was aimed to be fixed. Controlling technique is using the PI variation of the PID controlling technique. In order to determine the object definitely morphological applications were used and the noise filtered by the median filter. The coordinate information of the object which was aimed to be tracked was determined as the pixel value which was in the gravity center of the object in the frame received from the camera. If the object within the image received from the camera away from the center point, margin of error is applied to the PI algorithm as a input. All this given procedures were written with C programming under a single program and were used in Raspberry Pi development card which has ARM11 microprocessor architecture. For image processing, OpenCV libraries were used.

Detarmination of patellar T EMG response with the Deep Tendon Reflex grading and tapping force

In the study, Patellar T Reflex has been evaluated with tapping force measurement and Deep Tendon Reflex grading. Tapping force is measured with an electronic tendon hammer which is designed by us also can be used as a trigger system for determination of patellar T response. System was tested in Akdeniz University Medicine Faculty Neurology Department EMG Laboratory. In the study at least ten trace were obtained from 10 subjects and a total of 100 traces evaluations were performed. The experiment group consisted of people with no neurological disorders. The expected impact of the study is presenting the relationship between tendon reflexes that trigged with applied force to the tendon, deep tendon reflex grading and EMG response as quantitative rating. This led to further understanding of the important role that the reflex behavior in the neurological examination. Also, other departments that can use that study explained at results section.

Identification and grading of spasticity by using AdaBoost and k-NN techniques

Spasticity is a disorder that causes movement constraints in daily life and induced by upper neuron syndrome. Although spasticity is a well-known motor disorder, there are related diagnostic and assessment problems. The lack of common rating in assessment of spasticity causes significant differences between the interpretations of the specialists. In this study, it is aimed to conduct a highly accurate evaluation based on the Ashworth scale, which is one of the most accepted scales for assessment and grading of the spasticity. In this study, Electromyogram (EMG) signals obtained in Rectus Femoris and Biceps Femoris muscles are used to trigger Patela-T reflexes in spastic patients. 26 features are derived by calculating five features in both time and frequency domains from measured EMG signals and six features of Pendulum movement to recognize the normal person and patients with Ashworth 1-Ashworth 2 spasticity levels. Finally, spasticity is graded by using k-nearest neighbor (k-NN) algorithm and AdaBoost method. The performance of the proposed study is evaluated by calculating accuracy, sensitivity, precision and specificity criteria. Simulation results are demonstrated that the study has crucial importance to assess and grade the spasticity.

Geographical origin of imported and domestic teas (Camellia sinensis) from Turkey as determined by stable isotope signatures

ABSTRACT In this study, stable isotope signatures (δ13C, δ15N, and δD) of both tea leaves and tea infusions were investigated to identify the geographical origin of Turkish domestic and imported tea samples. Sixteen domestic tea samples collected from different locations in the Black Sea Region, which produces almost 100% of tea in Turkey, and 11 imported tea samples (Kenya, India, Sri Lanka, Indonesia, and China) purchased from importers were studied. δ13C, δ15N, and δD in the samples were determined using isotope ratio mass spectrometry (IR-MS). δ13C in the samples ranged from −29.18 ± 0.01 to −25.7 ± 0.2, while δ15N ranged between 1.1 ± 0.2 and 5.2 ± 0.8. However, δD in the samples were found to be in the range from 56.5 ± 0.3 to 72 ± 1. The classifications of the tea samples into domestic and imported tea samples were achieved with 100% accuracy using multivariate statistical analyses (principal component analysis, PCA, and hierarchical cluster analysis, HCA). In conclusion, the domestic tea samples had a distinctive isotopic fingerprint and the isotopic ratios used in the study can be significant predictors in determination of the geographical source of Turkish tea.

Extract gait pattern with Kinect

The prominence of systems for automatic person identification has risen increasingly during in recent years. One of the biometric features such as finger print, face, iris, hand and DNA is human gait which the only knows biometric feature distinguishes the other biometric from perceptibility some distance away. In former work, gait recognition is mainly achieved with camera systems. In this paper, we propose an approach for gait pattern based on ASUS Xtion Pro Live Kinect sensor with an integrated depth sensor alowing for skeleton detection and tracking in real time. For the extraction of gait pattern record are taken from segital plane and accuray of gait pattern are supported from real planes assays.

System design for elderly fall

Falls in the elderly, due to the anatomical and physiological changes that occur with the effects of aging is common in the elderly. These injuries can negatively affect the life of the elderly. Early intervention is vital in this fall accidents. Late intervened fall accidents result in death. In this study, we propose a fall detection system that allows the detection of fall accidents and emergency call for first aid. In this system, obtained accelerometer data from the wearable eZ430-Chronos watch, which is placed on the elderly individuals arm, will be transmitted wirelessly to the Raspberry PI mini computer. The obtained data will be analyzed with the designed fall detection algorithm. If the fall accident is detected, a notification will be sent using Google Cloud Messaging (GCM) services to smartphones of elderly relatives. With the designed system, fall accidents will immediately be determined and elderly relatives will be warned.

Indoor location service with beacons on embedded linux

The trend of using location-based services (LBS) is not new but has evolved over long time and continues to enter into new application and service landscapes. LBS have essentially targeted outdoor scenarios, whereas GPS is the de-facto standard for these cases. Indoor localization technologies are also gaining attraction. In this paper, we present our Linux-based solution, which runs on low-cost embedded hardware, and is equipped with shortrange wireless technologies such as Bluetooth and WIFI. Using the beaconing capabilities of these technologies in combination with control over M2M protocols, we are able provide different applications at the point of presence (PoP).

Area-wide ANPR coverage with a small number of instrumented vehicles

The paper investigates the applicability of an area-wide coverage of speed enforcement by a non-stationary Automatic Number Plate Recognition (ANPR) system. Two instrumented cars, each fitted with an ANPR camera and related equipment, were facilitated in the surveys. These cars were parked systematically according to a schedule during the day within a large university campus in Turkey on 11 corridors over a four-month period. The work was carried out at two stages to be able to compare the effectiveness of the application. During the first two months, the application was disguised from the drivers. In the next two-month period the public were informed of the scheme by various publicity means and were told to obey the speed limits. Since the tests were carried out within a closed area, in both stages more or less the same profile of drivers were studied, hence the findings are expected to reflect directly the effects of the application. It was found that in most corridors the average speeds were lower in the second stage. For the two month period after the announcement of the scheme, no evidence was found to show that the impact was diminishing, suggesting the effectiveness of the system. However, the study revealed the problem of posting unsuitable speed limits in the campus which requires further attention.

Assesment of spasticity via patellar pendulum and muscular-skeletal modelling

Spasticity is a movement disorder developing after upper motor neuron lesion. It is the tonus increase associated with movement velocity of extremity muscles, and thus it affects the daily lives of patients. It is quite a common disease as it develops due to lesions in brain and spinal cord. This study defines the electrophysiological characteristics of Patellar T reflex and patellar pendulum which triggered by patellar T reflex. Spasticity changes the responses by utilizing movement features of knee joint during the reflex and pendulum. Muscular-Skeletal simulation is employed to validate and visualize spasticity condition by calculating muscular forces and activation levels compared to EMG data. Also the study attempts to develop a measurement method to be used in spasticity.

Image-processing based determination of kinesiologic parameters of spasticity patients

In this study, it is aimed to obtain Patella T-Reflex kinesiologic parameters on spasticity patients by using image processing methods and thereby extracting objective data for the evaluation of spasticity patients. For this reason, the movement of Patella T-reflex response to knee joint was recorded with the help of a camera after three LED markers placed on the knee joint. For the correctness of the pattern of the movements, data were compared with two-axis goniometer of Biometric Ltd company and thereby it was shown that the reflex patterns are identical. Thereon, using the software we have developed, the Patella T-reflex kinesiologic parameters were calculated and data were obtained for objective evaluation of the Spasticity patients.