Murad Saleh Alfarzaeai

Dr. Murad Saleh Alfarzaeai

Information Institute, Ministry of Emergency Management of the People’s Republic of China, Beijing, China
orcid

Research Focus & Metrics

Academic Career Length

7Years (2020-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
24 April 2020 in IEEE Access
Verified
Journal Article
Open access
Murad Saleh AlfarzaeaiMurad Saleh Alfarzaeai
Murad Saleh Alfarzaeai
QNQiang Niu
Qiang Niu
JZJiaqi Zhao
Jiaqi Zhao
RERefat Eshaq
Refat Eshaq
EHEryi Hu
Eryi Hu
Recognition and separation of Coal/Gangue are important phases in the coal industries for many aspects. This paper addressed the topic of Coal/Gangue recognition and built a new model called (CGR-CNN) based on Convolutional Neural network (CNN) and using thermal images as standard images for Coal/Gangue recognition. The CGR-CNN model has been developed, augmentation principle has been applied in order to increase the dataset and the best experimental results have been achieved (99.36%) learning accuracy and (95.09%) validation accuracy, in the prediction phase (160) new images of coal and gangue (80 for both) have been tested to measure the efficiency of the work, the prediction result comes with (100%) for coal recognition accuracy and (97.5%) gangue recognition accuracy giving an overall prediction accuracy (98.75%).

17 March 2020 in IEEE Access
Verified
Journal Article
Open access
RERefat Eshaq
Refat Eshaq
EHEryi Hu
Eryi Hu
LMLi Menggang
Li Menggang
Murad Saleh AlfarzaeaiMurad Saleh Alfarzaeai
Murad Saleh Alfarzaeai
Distinguishing between coal and gangue in the production lines of mining factories based on the thermal energy and infrared radiation emission of an object is feasible. In this paper, we use an infrared camera (IC) to distinguish between coal and gangue in the industrial mining field. Additionally, this system is considered to be a binary classification system that has two classes. We analyze the infrared images of coal and gangue; then extract the appropriate texture features from the infrared images in order to develop an accurate classification system by using support vector machine (SVM). The method applied in this work essentially depends on feature extraction of images. The statistical features based on gray level information (GLI), grey-level cooccurrence matrix (GLCM) and visual features are executed. Thus, we suggest preparation steps to obtain one select feature before importing the data into the SVM classifier, and this approach is adopted as the fundamental basis for our work. We exploit only one feature of the infrared image, namely, Cb, which is extracted from the YCbCr color space, and then compute the mean value of Cb after heating and capturing the photos for the coal and gangue samples. The proposed method achieves a high classification accuracy 97.83 % by using Gaussian-SVM.

20 February 2023 in Energies
Verified
Journal Article
Open access
PDF file
Murad Saleh AlfarzaeaiMurad Saleh Alfarzaeai
Murad Saleh Alfarzaeai
EHEryi Hu
Eryi Hu
WPWang Peng
Wang Peng
Qiang NiuQiang Niu
Qiang Niu
MAMaged M. A. Alkainaeai
Maged M. A. Alkainaeai
Computer-vision-based separation methods for coal gangue face challenges due to the harsh environmental conditions in the mines, leading to the reduction of separation accuracy. So, rather than purely depending on the image features to distinguish the coal gangue, it is meaningful to utilize fixed coal characteristics like density. This study achieves the classification of coal and gangue based on their mass, volume, and weight. A dataset of volume, weight and 3_side images is collected. By using 3_side images of coal gangue, the visual perception value of the volume is extracted (ExM) to represent the volume of the object. A Support Vector Machine (SVM) classifier receives (ExM) and the weight to perform the coal gangue classification. The proposed system eliminates computer vision problems like light intensity, dust, and heterogeneous coal sources. The proposed model was tested with a collected dataset and achieved high recognition accuracy (KNN 100%, Linear SVM 100%, RBF SVM 100%, Gaussian Process 100%, Decision Tree 98%, Random Forest 100%, MLP 100%, AdaBosst 100%, Naive Bayes 98%, and QDA 99%). A cross-validation test has been done to verify the generalization ability. The results also demonstrate high classification accuracy (KNN 96%, Linear SVM 100%, RBF SVM 96%, Gaussian Process 96%, Decision Tree 99%, Random Forest 99%, MLP 100%, AdaBosst 99%, Naive Bayes 99%, and QDA 99%). The results show the high ability of the proposed technique ExM-SVM in coal gangue classification tasks.

1 January 2024 in Computers, Materials & Continua
Verified
Journal Article
Open access
PDF file
ZGZenglun Guan
Zenglun Guan
Murad Saleh AlfarzaeaiMurad Saleh Alfarzaeai
Murad Saleh Alfarzaeai
EHEryi Hu
Eryi Hu
TATaqiaden Alshmeri
Taqiaden Alshmeri
WPWang Peng
Wang Peng
In the coal mining industry, the gangue separation phase imposes a key challenge due to the high visual similarity between coal and gangue. Recently, separation methods have become more intelligent and efficient, using new technologies and applying different features for recognition. One such method exploits the difference in substance density, leading to excellent coal/gangue recognition. Therefore, this study uses density differences to distinguish coal from gangue by performing volume prediction on the samples. Our training samples maintain a record of 3-side images as input, volume, and weight as the ground truth for the classification. The prediction process relies on a Convolutional neural network (CGVP-CNN) model that receives an input of a 3-side image and then extracts the needed features to estimate an approximation for the volume. The classification was comparatively performed via ten different classifiers, namely, K-Nearest Neighbors (KNN), Linear Support Vector Machines (Linear SVM), Radial Basis Function (RBF) SVM, Gaussian Process, Decision Tree, Random Forest, Multi-Layer Perceptron (MLP), Adaptive Boosting (AdaBosst), Naive Bayes, and Quadratic Discriminant Analysis (QDA). After several experiments on testing and training data, results yield a classification accuracy of 100%, 92%, 95%, 96%, 100%, 100%, 100%, 96%, 81%, and 92%, respectively. The test reveals the best timing with KNN, which maintained an accuracy level of 100%. Assessing the model generalization capability to new data is essential to ensure the efficiency of the model, so by applying a cross-validation experiment, the model generalization was measured. The used dataset was isolated based on the volume values to ensure the model generalization not only on new images of the same volume but with a volume outside the trained range. Then, the predicted volume values were passed to the classifiers group, where classification reported accuracy was found to be (100%, 100%, 100%, 98%, 88%, 87%, 100%, 87%, 97%, 100%), respectively. Although obtaining a classification with high accuracy is the main motive, this work has a remarkable reduction in the data preprocessing time compared to related works. The CGVP-CNN model managed to reduce the data preprocessing time of previous works to 0.017 s while maintaining high classification accuracy using the estimated volume value.

19 November 2024 in IEEE Robotics and Automation Letters
Verified
Journal Article
Closed access
TATaqiaden Alshameri
Taqiaden Alshameri
PWPeng Wang
Peng Wang
DLDaheng Li
Daheng Li
WWWei Wei
Wei Wei
HDHaonan Duan
Haonan Duan
Murad Saleh AlfarzaeaiMurad Saleh Alfarzaeai
Murad Saleh Alfarzaeai
Grasp is at the core of robotic manipulation tasks. Nonetheless, most 6-DOF methods resort to a one-time setup via intensive analytics and targeting a predetermined domain. On the other hand, learning and adapting in real environments is of great promise to robotics yet challenging. In this context, this letter presents a grasp learning agent embodied in a dual-arm robot with a gripper and suction ends. Theoretically, GraspAgent is equipped with two adversarial components with the capacity for continual improvement. The first component aims to learn the synthesis of high dexterity grasps using a novel Grasp Awareness Generative Adversarial Network . Thanks to the designed set of contrastive objectives, GA-GAN yields a high rate of feasible grasps for clutter scenes. The second component is an Adversarial Experience Replay . AER promotes certain attributes during the regular rounds of training. For instance, a quality network is forced to adapt the maturity of the sampler by highlighting regions where the sampler performs well. Further, the quality network of each arm is encouraged to learn different grasp primitives, which reflects on better learning capacity. Finally, Experiments on a Yumi robot reveal a final average success rate of 93% after exploiting 17k+ feedback data. Moreover, We show that grasp sampling with GA-GAN surpasses a set of recent baseline methods.