Norasmadi Abdul Rahim

Moving vehicle noise classification using backpropagation algorithm

The hearing impaired is afraid of walking along a street and living a life alone. Since, it is difficult for hearing impaired to hear and judge sound information and they often encounter risky situations while they are in outdoors. The sound produced by moving vehicle in outdoor situation cannot be moderate wisely by profoundly deaf people. They also cannot distinguish the type and the distance of any moving vehicle approaching from their behind. Generally the profoundly deaf people do not use any hearing aid which does not provide any benefit. In this paper, a simple system that identifies the type and distance of a moving vehicle using artificial neural network has been proposed. The noises emanated from moving vehicles along the roadside were recorded along with the type and distance of moving vehicles. Simple feature extraction algorithm for extracting the feature from noise emanated by the moving vehicle has been made using frequency analysis approach. A one-third-octave filter bands is used for getting the important signatures from the emanated noise. The extracted features are associated with the type and distance of the moving vehicle and a simple neural network model is developed. The developed neural network model is tested for its validity.

In-vitro diagnosis of single and poly microbial species targeted for diabetic foot infection using e-nose technology

BackgroundEffective management of patients with diabetic foot infection is a crucial concern. A delay in prescribing appropriate antimicrobial agent can lead to amputation or life threatening complications. Thus, this electronic nose (e-nose) technique will provide a diagnostic tool that will allow for rapid and accurate identification of a pathogen.ResultsThis study investigates the performance of e-nose technique performing direct measurement of static headspace with algorithm and data interpretations which was validated by Headspace SPME-GC-MS, to determine the causative bacteria responsible for diabetic foot infection. The study was proposed to complement the wound swabbing method for bacterial culture and to serve as a rapid screening tool for bacteria species identification. The investigation focused on both single and poly microbial subjected to different agar media cultures. A multi-class technique was applied including statistical approaches such as Support Vector Machine (SVM), K Nearest Neighbor (KNN), Linear Discriminant Analysis (LDA) as well as neural networks called Probability Neural Network (PNN). Most of classifiers successfully identified poly and single microbial species with up to 90% accuracy.ConclusionsThe results obtained from this study showed that the e-nose was able to identify and differentiate between poly and single microbial species comparable to the conventional clinical technique. It also indicates that even though poly and single bacterial species in different agar solution emit different headspace volatiles, they can still be discriminated and identified using multivariate techniques.

A mapping mobile robot using RP Lidar scanner

The environment mapping in one of the necessary aspects in mobile robotics studies when dealing with localization, positioning, autonomous navigation, as well as search and rescue. Its success depends on the accuracy and reliability of its implementation and may depend on sensors which are used to acquire the environment data. This paper presents work on mapping which uses a low cost laser rangefinder, developed by Robopeak, RP Lidar. This work will be implemented on an autonomous mobile robot for indoor environment applications. Initially, a mobile robot incorporated with the sensor was developed after which a wireless monitoring and control station was established in order to perform data collection of the environments. A research laboratory with tables and equipment is used as a testbed. The collected data were analysed using three methods of pre-processing; i. raw filter, ii. a moving average of smooth filter, and iii. the combination of the raw filter and the moving average. The results show the environment map based on the raw data and the pre-processing performances. The combination of the raw filter and the moving average performs the best result. Besides, the scanning accuracy of the developed system is successful with more than 90% of correctness. As a conclusion, the laser rangefinder and the pre-processing used is capable to map the environment with clear image and hence can be used for a variety of applications.

Improving performance of 2D SLAM methods by complementing Kinect with laser scanner

The feasibility of using Kinect sensor for 2D Simultaneous Localization and Mapping (SLAM) application has been widely studied. Researchers concluded that the acquired maps are often inaccurate due to the limited field of view of the sensor. Therefore in this work, we complemented the Kinect with a laser scanner and proposed a method to merge the data from both sensors. Two SLAM algorithms (i.e Gmapping and Hector SLAM) were tested using the method, in different environments. The results show that the method is able to detect multi-sized objects and produce more accurate map as compared to when using single sensor (i.e Kinect only or laser scanner only). Finally, the performance of the Gmapping and Hector SLAM are compared particularly in terms of the computational complexity and the map accuracy.

Classification of Malaysia aromatic rice using multivariate statistical analysis

Aromatic rice (Oryza sativa L.) is considered as the best quality premium rice. The varieties are preferred by consumers because of its preference criteria such as shape, colour, distinctive aroma and flavour. The price of aromatic rice is higher than ordinary rice due to its special needed growth condition for instance specific climate and soil. Presently, the aromatic rice quality is identified by using its key elements and isotopic variables. The rice can also be classified via Gas Chromatography Mass Spectrometry (GC-MS) or human sensory panels. However, the uses of human sensory panels have significant drawbacks such as lengthy training time, and prone to fatigue as the number of sample increased and inconsistent. The GC–MS analysis techniques on the other hand, require detailed procedures, lengthy analysis and quite costly. This paper presents the application of in-house developed Electronic Nose (e-nose) to classify new aromatic rice varieties. The e-nose is used to classify the variety of aromatic...

Identification of DC motor drive system model using Radial Basis Function (RBF) Neural Network

In this paper, we present a Radial Basis Function Neural Network (RBFNN)-based Nonlinear Auto-Regressive Model with Exegeneous Inputs (NARX) model of a DC motor drive controller model by (Rahim, 2004). Tests were conducted to measure the accuracy of the model (using One Step Ahead (OSA) and its validity (using correlation tests and histogram analysis). The resulting model produced Mean Square Error (MSE) of 8.53 × 10−3 and 8.82 × 10−3 on the training set and test set, respectively, while fulfilling all validation tests performed.

Classification of human breathing task based on electromyography signal of respiratory muscles

Breathing is one of the human physiological activities that catch the interest of researchers especially in the area of medical diagnosis and human physiological performance. Apart from conventional measurement using intake or outflow of air, breathing characteristics could also be assessed through human respiratory muscles with the analysis on Electromyography (EMG) signal. In this paper, EMG signal of human breathing is acquired from four respiratory muscles i.e. sternocleidomastoid, scalene, intercostal muscle and diaphragm while subjects perform four different breathing tasks. The aim is to classify EMG features from the muscles into the four breathing tasks. Classification is done using Feedforward Multi-layer Perceptron Artificial Neural Network (MLPANN). Four features are derived from the EMG data i.e. root-mean-square (RMS), zero crossing (ZC), mean frequency (MNF) and mean frequency power (MP). Classification is performed to compare the accuracy result of input vector from the four features of EMG and three combination set of these features using i) four data segmentation frame sizes and ii) six number of hidden neurons. The result of data classification shows highest accuracy when all feature sets is used as input to MLPANN with segmentation frame size of 1000 ms and number of hidden neurons of 60. Classifation accuracy obtained is 59.52%.

Laser range finder filters for mapping and localisation

In the real world, sensors are always exposes to noise and error during measurements. These problems will lead to invalid and overshoot readings, as well as blank data. Fortunately, these errors can be treated using filters. For data collection process using laser range finder, the use of filters is necessary as they can help to overcome the laser limitations and improve the scanning input. Here, this paper presents a significant work on filters towards the laser range finder data for robotic mapping and localisation purpose. Raw filter and smooth filter are chosen and they will assist the localisation algorithm adapted here in determining the accurate coordinates of mobile robot location estimation. The results show that by using the filters, location estimation have achieved 90% accuracy. At the same time, it is also shown that less processing data and faster processing speed can be achieved by incorporating the filters. As a conclusion, the raw and smooth filters are very suitable to be applied into the laser range finder data and are successful in improving the localisation accuracy.

Athlete Overtraining Monitoring System

This paper introduces the design and development of a system that can early detect overtraining problem during training activities. These problems can affect athletes’ physiological and psychological conditions as well as reducing their performance. Maximum heart rate (MHR) is the limitation heart rate of athlete and can be indicator for overtraining exercise. Heart rate is usually used to detect and prevent overtraining by coaches and athlete. In this study, we use electrocardiograph (ECG) for amplifying and filtering the signal from the body. National instrument (NI) DAQ is used to acquire the real-time signal from the sensors circuit and pass the data to LabVIEW for real-time monitoring and analysis. The software displays heart rate as well as detecting the abnormality if present. Furthermore, it also features a simple yet comprehensive user interface where the athlete data, date and time for the data collection are saved in the specified txt file for future reference.