Zuliani Zulkoffli
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Research Interests
Zuliani Zulkoffli has a total of 16 co-authors. Here are the top co-authors by number of shared publications:
Elmi Abu Bakar
University Science Malaysia
3 shared publications- Huang Shen Chua
School of Engineering, University of Wollongong Malaysia, Glenmarie Campus, Shah Alam, Malaysia
1 shared publication - Jiehan Teoh
Department of Computing and Information Systems, School of Engineering, Research Centre for Human-Machine Collaboration, Sunway University, Petaling Jaya, Malaysia
1 shared publication - Kian Meng Yap
Research Centre for Human-Machine Collaboration, School of Computing and Artificial Intelligence, Faculty of Engineering and Technology, Sunway University, Petaling Jaya, Malaysia
1 shared publication Kian-Yiap Lim
1 shared publicationKumaraguru Rengasamy
1 shared publication
Latest Publications
- Open Access
- 15 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 173686-173699
Exploring Generative AI for YOLO-Based Object Detection to Enhance Flood Disaster Response in Malaysia
Flood disasters pose significant threats to communities around the world. In Malaysia, floods are recurrent and severe natural disasters due to its geographical location. Focusing on the response phase of flood disaster management, this study identifies critical issues including the limitations of manual observation in traditional methods and the lack of structured flood image databases. Hence, this research explores the use of generative artificial intelligence (gen-AI) to create synthetic images as an alternative source in flood disaster detection and response. By adopting a systematic approach with the You Only Look Once (YOLO) algorithm, the model is trained on gen-AI produced synthetic images and then tested using real drone-captured images from past flood events. The model demonstrates promising preliminary results with the highest performing model achieving a mean Average Precision (mAP50) of 96.7%, mAP @[0.5,0.95] of 78.7% and a F1-score of 90.8%. These findings highlight the potential of gen-AI in supporting flood disaster response efforts in Malaysia, aiming to integrate the mix of synthetic images during non-flooding periods and actual drone images during flood scenarios into a comprehensive framework, FloodIntel, for future research.
- 15 November 2024
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 12, 173686-173699
Open AccessExploring Generative AI for YOLO-Based Object Detection to Enhance Flood Disaster Response in Malaysia
Flood disasters pose significant threats to communities around the world. In Malaysia, floods are recurrent and severe natural disasters due to its geographical location. Focusing on the response phase of flood disaster management, this study identifies critical issues including the limitations of manual observation in traditional methods and the lack of structured flood image databases. Hence, this research explores the use of generative artificial intelligence (gen-AI) to create synthetic images as an alternative source in flood disaster detection and response. By adopting a systematic approach with the You Only Look Once (YOLO) algorithm, the model is trained on gen-AI produced synthetic images and then tested using real drone-captured images from past flood events. The model demonstrates promising preliminary results with the highest performing model achieving a mean Average Precision (mAP50) of 96.7%, mAP @[0.5,0.95] of 78.7% and a F1-score of 90.8%. These findings highlight the potential of gen-AI in supporting flood disaster response efforts in Malaysia, aiming to integrate the mix of synthetic images during non-flooding periods and actual drone images during flood scenarios into a comprehensive framework, FloodIntel, for future research.
- 11 August 2022
- Published by Springer Nature
- p. 101-112
Drowsiness Detection for Safe Driving Using PERCLOS and YOLOv2 Method
- 11 August 2022
- Published by Springer Nature
- p. 101-112
Drowsiness Detection for Safe Driving Using PERCLOS and YOLOv2 Method
- Open Access
- 1 May 2020
- Published by Institute of Advanced Engineering and Science in Indonesian Journal of Electrical Engineering and Computer Science
- Vol. 18 (2) , 790-798
Partial pose estimation of 3D rigid object system using outer box method
This article introduces a novel approach for identify partial pose estimation using template matching method. The algorithms performs 2D correlation matching on tested image to CAD database by using regional shape representation in order to get the similar object pose in CAD database. The descriptor named outer box method, it is useful for rescale or aligning object size in both different images of tested image and CAD database image and also provide interest point for segmentation in image registration stage. The proposed algorithm were experimentally shown to be robust to apply on scale changes, various complex shape, unstructured CAD database and mixed CAD model database. Last part, the identified pose and its retrieved pose angle was calculated and achieved high accuracyin range ±0.388˚ to ±1.471˚.
- 1 May 2020
- Published by Institute of Advanced Engineering and Science in Indonesian Journal of Electrical Engineering and Computer Science
- Vol. 18 (2) , 790-798
Open AccessPartial pose estimation of 3D rigid object system using outer box method
This article introduces a novel approach for identify partial pose estimation using template matching method. The algorithms performs 2D correlation matching on tested image to CAD database by using regional shape representation in order to get the similar object pose in CAD database. The descriptor named outer box method, it is useful for rescale or aligning object size in both different images of tested image and CAD database image and also provide interest point for segmentation in image registration stage. The proposed algorithm were experimentally shown to be robust to apply on scale changes, various complex shape, unstructured CAD database and mixed CAD model database. Last part, the identified pose and its retrieved pose angle was calculated and achieved high accuracyin range ±0.388˚ to ±1.471˚.
- 1 August 2019
- Published by Elsevier in Journal of Environmental Chemical Engineering
- Vol. 7 (4)
Comparison of activated carbon and zeolites’ filtering efficiency in freshwater
- 1 August 2019
- Published by Elsevier in Journal of Environmental Chemical Engineering
- Vol. 7 (4)
Comparison of activated carbon and zeolites’ filtering efficiency in freshwater
- 1 June 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Detection of Brain Tumor and Extraction of Features in MRI Images Using K-means Clustering and Morphological Operations
In the past couple of decades, progressive research has been done in the medical imaging field to accurately detect brain tumors. This is vital for early treatment and saves time. This report presents a method of automatic detection of brain tumor in magnetic resonance imaging (MRI) images. The proposed algorithm consists of three stages: image acquisition, pre-processing and post-processing. In image acquisition, a database of brain tumor MRI images is developed with the help of publicly available online sources. The pre-processing stage involves converting the image to grayscale and applying noise filters to enhance the image. In the post-processing stage, k-means clustering algorithm is used to separate the image into four clusters of varying intensity levels. The cluster with the tumor is found and then the tumor is extracted by using morphological and region properties operations. Features like energy, contrast, kurtosis, correlation and homogeneity are found along with the area and perimeter of the tumor and the accuracy of the extraction. These extracted features can be used in the future to more accurately classify brain tumors with the help of neural networks. The proposed method has given good results and accuracy.
- 1 June 2019
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Detection of Brain Tumor and Extraction of Features in MRI Images Using K-means Clustering and Morphological Operations
In the past couple of decades, progressive research has been done in the medical imaging field to accurately detect brain tumors. This is vital for early treatment and saves time. This report presents a method of automatic detection of brain tumor in magnetic resonance imaging (MRI) images. The proposed algorithm consists of three stages: image acquisition, pre-processing and post-processing. In image acquisition, a database of brain tumor MRI images is developed with the help of publicly available online sources. The pre-processing stage involves converting the image to grayscale and applying noise filters to enhance the image. In the post-processing stage, k-means clustering algorithm is used to separate the image into four clusters of varying intensity levels. The cluster with the tumor is found and then the tumor is extracted by using morphological and region properties operations. Features like energy, contrast, kurtosis, correlation and homogeneity are found along with the area and perimeter of the tumor and the accuracy of the extraction. These extracted features can be used in the future to more accurately classify brain tumors with the help of neural networks. The proposed method has given good results and accuracy.
- Open Access
- 1 February 2016
- Published by IOP Publishing in IOP Conference Series: Materials Science and Engineering
- Vol. 114 (1) , 012038
Pose Estimation using 1D Fourier Transform and Euclidean Distance Matching of CAD Model and Inspected Model Part
This paper present pose estimation relation of CAD model object and Projection Real Object (PRI). Image sequence of PRI and CAD model rotate on z axis at 10 degree interval in simulation and real scene used in this experiment. All this image is go through preprocessing stage to rescale object size and image size and transform all the image into silhouette. Correlation of CAD and PRI image is going through in this stage. Magnitude spectrum shows a reliable value in range 0.99 to 1.00 and Phase spectrum correlation shows a fluctuate graph in range 0.56 - 0.97. Euclidean distance correlation graph for CAD and PRI shows 2 zone of similar value due to almost symmetrical object shape. Processing stage of retrieval inspected PRI image in CAD database was carried out using range phase spectrum and maximum magnitude spectrum value within ±10% tolerance. Additional processing stage of retrieval inspected PRI image using Euclidean distance within ±5% tolerance also carried out. Euclidean matching shows a reliable result compared to range phase spectrum and maximum magnitude spectrum value by sacrificing more than 5 times processing time.
- 1 February 2016
- Published by IOP Publishing in IOP Conference Series: Materials Science and Engineering
- Vol. 114 (1) , 012038
Open AccessPose Estimation using 1D Fourier Transform and Euclidean Distance Matching of CAD Model and Inspected Model Part
This paper present pose estimation relation of CAD model object and Projection Real Object (PRI). Image sequence of PRI and CAD model rotate on z axis at 10 degree interval in simulation and real scene used in this experiment. All this image is go through preprocessing stage to rescale object size and image size and transform all the image into silhouette. Correlation of CAD and PRI image is going through in this stage. Magnitude spectrum shows a reliable value in range 0.99 to 1.00 and Phase spectrum correlation shows a fluctuate graph in range 0.56 - 0.97. Euclidean distance correlation graph for CAD and PRI shows 2 zone of similar value due to almost symmetrical object shape. Processing stage of retrieval inspected PRI image in CAD database was carried out using range phase spectrum and maximum magnitude spectrum value within ±10% tolerance. Additional processing stage of retrieval inspected PRI image using Euclidean distance within ±5% tolerance also carried out. Euclidean matching shows a reliable result compared to range phase spectrum and maximum magnitude spectrum value by sacrificing more than 5 times processing time.
- 22 December 2015
- Published by Penerbit UTM Press in Jurnal Teknologi (Sciences & Engineering)
- Vol. 78 (1)
TEMPLATE BASED DEFECT DETECTION OF FLEXIBLE PRINTED CIRCUIT
This study presented a research on machine vision inspection to define defects on flexible printed circuit (FPC). The images were subjected to image processing system where an elimination-subtraction method used. In this algorithm, 7 types of FPC defects defined and simulated in the system using Specimen 1 and processing time taken for both side inspection was 3.3s. Then, the commercial patent design of FPC was tested as specimen 2 to define short circuit defects on it. The processing time taken by this algorithm on specimen 2 was 0.28s. Comparison on manual inspection and machine vision implementation were carried out and greatly resulted on shorten inspection time to 59.7%. This result shows significant contribution in increasing the efficiency of FPC inspection process.
- 22 December 2015
- Published by Penerbit UTM Press in Jurnal Teknologi (Sciences & Engineering)
- Vol. 78 (1)
TEMPLATE BASED DEFECT DETECTION OF FLEXIBLE PRINTED CIRCUIT
This study presented a research on machine vision inspection to define defects on flexible printed circuit (FPC). The images were subjected to image processing system where an elimination-subtraction method used. In this algorithm, 7 types of FPC defects defined and simulated in the system using Specimen 1 and processing time taken for both side inspection was 3.3s. Then, the commercial patent design of FPC was tested as specimen 2 to define short circuit defects on it. The processing time taken by this algorithm on specimen 2 was 0.28s. Comparison on manual inspection and machine vision implementation were carried out and greatly resulted on shorten inspection time to 59.7%. This result shows significant contribution in increasing the efficiency of FPC inspection process.