Browsing by Author "Bulugu, Isack"

Now showing 1 - 10 of 10
  • Thumbnail Image
    Item
    Algorithm for License Plate Localization and Recognition for Tanzania Car Plate Numbers
    (International Journal of Science and Research, 2013-05) Bulugu, Isack
    In this paper, License plate localization and recognition (LPLR) is presented. It uses image processing and character recognition technology in order to identify the license number plates of the vehicles automatically. This system is considerable interest because of its good application in traffic monitoring systems, surveillance devices and all kind of intelligent transport system. The objective of this work is to design algorithm for License Plate Localization and Recognition (LPLR) of Tanzanian License Plates. The plate numbers used are standard ones with black and yellow or black and white colors. Also, the letters and numbers are placed in the same row (identical vertical levels), resulting in frequent changes in the horizontal intensity. Due to that, the horizontal changes of the intensity have been easily detected, since the rows that contain the number plates are expected to exhibit many sharp variations. Hence, the edge finding method is exploited to find the location of the plate. To increase readability of the plate number, part of the image was enhanced, noise removal and smoothing median filter is used due to easy development. The algorithm described in this paper is implemented using MATLAB 7.11.0(R2010b).
  • Thumbnail Image
    Item
    Deep Predictive Neural Network: Unsupervised Learning for Hand Pose Estimation
    (2019-08-15) Banzi, Jamal; Bulugu, Isack; Ye, Zhongfu
    The discriminative approaches for hand pose estimation from depth images usually require dense annotated data to train a supervised network. Additionally, generative methods depend on temporal information in generating candidate poses which can be trapped due to local minima during the optimization process. Different from these methods, we propose a hybrid two-stage deep predictive neural network approach that performs predictive coding of image sequences of hand poses in order to capture latent features underlying a given image. Firstly, we train a deep convolutional neural network (CNN) for direct regression of hand joints position. Secondly, we add an unsupervised error term as a part of the recurrent architecture connected with predictive coding portion. An error regression term (ERT) ensures minimal residual errors of the estimated values while the predictive coding portion allows training of the network without the supervision of image sequences, so no dense annotation of data is required. We conduct a complete experiment using two challenging public datasets, ICVL and NYU. Using the ICVL datasets, our approach improved accuracy over the current state of the art methods with an average error joint of 7.5mm. We also achieve 12.2mm average error joint on NYU dataset which is the smallest error to be achieved on all state-of-art approaches.
  • Thumbnail Image
    Item
    Higher-Order Local Autocorrelation Feature Extraction Methodology for Hand Gestures Recognition
    (IEEE, 2017-12-25) Bulugu, Isack; Ye, Zhongfu; Banzi, Jamal; Bulugu, Isack
    A novel feature extraction method for hand gesture recognition from sequences of image frames is described and tested. The proposed method employs higher order local autocorrelation (HLAC) features for feature extraction. The features are extracted using different masks from Grey-scale images for characterising hands image texture with respect to the possible position, and the product of the pixels marked in white. Then features with the most useful information are selected based on mutual information quotient (MIQ). Multiple linear discriminant analysis (LDA) classifier is adopted to classify different hand gestures. Experiments on the NUS dataset illustrate that the HLAC is efficient for hand gesture recognition compared with other feature extraction methods.
  • Thumbnail Image
    Item
    An Improved Method for Tanzania Number Plate Location and Segmentation Based on Mathematical Morphology and Regional Features of an Image
    (International Journal of Science and Research, 2013-12) Bulugu, Isack; Zhijun, Pei
    In the Automatic Number Plate Recognition System (ANPR), Plate Number Location and Character segmentation are very important parts of an ANPR system before Recognition part. In this paper, plate number localization and character segmentation using mathematical morphological approach and regional features of images are discussed from the proposed ANPR system of vehicles in Tanzania. The proposed algorithm consists of three main modules: Pre-processing (cutting and resizing, convert RGB image to grayscale image, image binarization use Otsu method), Finding Region of Interest(morphology opening to remove noises & dilation operation, measure properties of image regions to find candidates), License plate exactly location (finding the LP angle & rotating LP, cut exactly LP region). The Character segmentation also consist three parts: Eliminate incrimination of the binary using boundary features, removes impurities using regional features and morphological process, and divide character into sub-images. The results show an average of 98% successful plate number localization and segmentation for proposed ANPR system in a total of 200 images captured from a complex outdoor environment in Tanzania. Implementation was done using MATLAB Version 7.11.0 (R2010b).
  • Thumbnail Image
    Item
    Learning a deep predictive coding network for a semi-supervised 3D-hand pose estimation
    (IEEE, 2020-03-27) Bulugu, Isack; Banzi, Jamal; Huang, Shiliang; Ye, Zhongfu
    In this paper we present a CNN based approach for a real time 3D-hand pose estimation from the depth sequence. Prior discriminative approaches have achieved remarkable success but are facing two main challenges: Firstly, the methods are fully supervised hence require large numbers of annotated training data to extract the dynamic information from a hand representation. Secondly, unreliable hand detectors based on strong assumptions or a weak detector which often fail in several situations like complex environment and multiple hands. In contrast to these methods, this paper presents an approach that can be considered as semi-supervised by performing predictive coding of image sequences of hand poses in order to capture latent features underlying a given image without supervision. The hand is modelled using a novel latent tree dependency model ( LDTM ) which transforms internal joint location to an explicit representation. Then the modeled hand topology is integrated with the pose estimator using data dependent method to jointly learn latent variables of the posterior pose appearance and the pose configuration respectively. Finally, an unsupervised error term which is a part of the recurrent architecture ensures smooth estimations of the final pose. Experiments on three challenging public datasets, ICVL, MSRA, and NYU demonstrate the significant performance of the proposed method which is comparable or better than state-of-the-art approaches.
  • Thumbnail Image
    Item
    Learning Hand Latent Features For Unsupervised 3D Hand Pose Estimation
    (2019-05-06) Banzi, Jamal; Bulugu, Isack
    Recent hand pose estimation methods require large numbers of annotated training data to extract the dynamic information from a hand representation. Nevertheless, precise and dense annotation on the real data is difficult to come by and the amount of information passed to the training algorithm is significantly higher. This paper presents an approach to developing a hand pose estimation system which can accurately regress a 3D pose in an unsupervised manner. The whole process is performed in three stages. Firstly, the hand is modelled by a novel latent tree dependency model (LTDM) which transforms internal joints location to an explicit representation. Secondly, we perform predictive coding of image sequences of hand poses in order to capture latent features underlying a given image without supervision. A mapping is then performed between an image depth and a generated representation. Thirdly, the hand joints are regressed using convolutional neural networks to finally estimate the latent pose given some depth map. Finally, an unsupervised error term which is a part of the recurrent architecture ensures smooth estimations of the final pose. To demonstrate the performance of the proposed system, a complete experiment is conducted on three challenging public datasets, ICVL, MSRA, and NYU. The empirical results show the significant performance of our method which is comparable or better than state-of-the-art approaches.
  • Thumbnail Image
    Item
    A novel Hand Pose Estimation using Dicriminative Deep Model and Transductive Learning Approach for Occlusion Handling and Reduced Descrepancy
    (IEEE, 2017-05-11) Banzi, Jamal; Bulugu, Isack; Ye, Zhongfu
    Discriminative based model have demonstrated an epic distinction in hand pose estimation. However there are key challenges to be solved on how to intergrate the self-similar parts of fingers which often occlude each other and how to reduce descrepancy among synthetic and realistic data for an accurate estimation. To handle occlusion which lead to inaccurate estimation, this paper presents a probabilistic model for finger position detection framework. In this framework the visibility correlation among fingers aid in predicting the occluded part between fingers thereby estimating hand pose accurately. Unlike convectional occlusion handling approach which assumes occluded parts of fingers as independent detection target, this paper presents a discriminative deep model which learns the visibility relationship among the occluded parts of fingers at multiple layers. In addition, we propose the semi-supervised Transductive Regression(STR) forest for classification and regression to minimise discrepancy among realistic and synthetic pose data. Experimental results demonstrate promising performance with respect to occlusion handling, and discrepancy reduction with higher degree of accuracy over state-of-the-art approaches.
  • Thumbnail Image
    Item
    Scale Invariant Static Hand-postures Detection using Extended Higher-order Local Autocorrelation Features
    (Foundation of Computer Science, 2016-02) Bulugu, Isack; Ye, ZhongFu
    This paper presents scale invariant static hand postures detection methods using extended HLAC features extractedfrom Log-Polar images. Scale changes of a handposture in an image are represented as shift in Log-Polar image. Robustness of the method is achieved through extracting spectral features from theeach row of the Log-Polar image. Linear Discriminant Analysis was used to combine features with simple classification methods in order to realize scale invariant hand postures detection and classification.The method was successful tested by performing experiment using NSU hand posture dataset images which consists 10 classes of postures, 24 samples of images per class, which are captured by the position and size of the hand within the image frame. The results showed that the detection rate using ExtendedHLAC can averaged reach 94.63% higher than using HLAC features on a Intel Core i5-4590 CPU running at 3.3 GHz.
  • Thumbnail Image
    Item
    Scale Invariant Static Hand-Postures Detection using Extended Higher-order Local Autocorrelation Features
    (Foundation of Computer Science (FCS), NY, USA, 2016-02-17) Bulugu, Isack; Ye, Zhongfu
    This paper presents scale invariant static hand postures detection methods using extended HLAC features extractedfrom Log-Polar images. Scale changes of a handposture in an image are represented as shift in Log-Polar image. Robustness of the method is achieved through extracting spectral features from theeach row of the Log-Polar image. Linear Discriminant Analysis was used to combine features with simple classification methods in order to realize scale invariant hand postures detection and classification.The method was successful tested by performing experiment using NSU hand posture dataset images which consists 10 classes of postures, 24 samples of images per class, which are captured by the position and size of the hand within the image frame. The results showed that the detection rate using Extended-HLAC can averaged reach 94.63% higher than using HLAC features on a Intel Core i5-4590 CPU running at 3.3 GHz.
  • Thumbnail Image
    Item
    Web-Accessible Liquid Distillation Column Temperature Monitoring and Control using ilab Shared Architecture
    (2013) Bulugu, Isack; Mwambela, Alfred
    The liquid distillation column is a research facility used in chemical and processing engineering at University of Dar es salaam. Originally it has been manually operated require the presence of operators all-time. It is used to control and take measurements data under hazardous operating condition. In this work the research equipment has been automated and its control and monitoring can now operate online, hence shared by many researchers. In this paper, the major findings were data acquisition device (DAQ) NI USB 6211 for acquiring temperatures and generate signals for control purposes, and thermocouples type K sensors were used since they can operate at very wide temperature range of -270 to 1370. The main parameter which has been monitored is temperature at various parts of the distillation column including temperature of the top/bottom and reboiler temperature. The approach used was computer controlled using Data acquisition device and labVIEW graphical programming language as standalone control system to extent it control and monitoring temperature online. The Interactive ilab shared architecture has been used to make the system accessible remotely. Labserver was created then interfaced with distillation column using NI USB 6211 as a data acquisition device. The DAQ was configured for temperature acquisition and signal generation for monitoring temperature and flux flow control respectively. For the remotely operation, the Labserver was connected to the LAN/internet and the distillation column was accessed through web services.