Analysis Of Personal Protective Equipment Classification Method Using Deep Learning
Personal Protective Equipment (PPE) is equipment worn in the workplace in order to reduce the exposure to hazards that causing serious injuries and illness to people. In this project, there are five types of PPE to be considered which is also recognized by Occupational Safety and Health Administrati...
| Main Author: | |
|---|---|
| Format: | Undergraduates Project Papers |
| Language: | English |
| Published: |
2022
|
| Subjects: | |
| Online Access: | http://umpir.ump.edu.my/id/eprint/39896/1/EA18087-SITI%20ZAHRAH%20NUR%20AIN-THESIS%20-%20Siti%20Zahrah.pdf |
| _version_ | 1825815382562177024 |
|---|---|
| author | Siti Zahrah Nur Ain, Silopung |
| author_facet | Siti Zahrah Nur Ain, Silopung |
| author_sort | Siti Zahrah Nur Ain, Silopung |
| collection | UMP |
| description | Personal Protective Equipment (PPE) is equipment worn in the workplace in order to reduce the exposure to hazards that causing serious injuries and illness to people. In this project, there are five types of PPE to be considered which is also recognized by Occupational Safety and Health Administration (OSHA) such as face mask, face shield, safety goggle, safety helmet and safety jacket. To avoid a tedious work in manually checking whether workers wear PPE or not, an automatic PPE classifier is constructed by utilizing a deep learning algorithm called Convolutional Neural Network (CNN). This classification is performed using Anaconda and Jupyter Notebok Software that use Python as the programming language. There are total of 7500 images in the PPE dataset, 6000 images for training with and another 1500 images for testing. The classification is done in two ways, one is by classifying the PPE into 5 classes and another one is by classifying into binary class after the best combined parameters are obtained using multiple training and testing by changing the parameters such as epoch, activation function, optimizer and filter layer. By using classifying into 5 classes, the final training accuracy is 89.39% and testing accuracy of 62.23%. On the other hand, by classifying the PPE into binary class, the PPE has final testing accuracy up 88% for all PPE. Face mask has final accuracy of 95.60%, face shield 94.32%, safety goggle 89.79%, safety helmet 98.90% and lastly safety jacket has 88.45% testing accuracy. Based on the result, CNN algorithm is a good algorithm as the binary classification of PPE achieved high accuracy result. |
| first_indexed | 2024-03-06T13:12:40Z |
| format | Undergraduates Project Papers |
| id | UMPir39896 |
| institution | Universiti Malaysia Pahang |
| language | English |
| last_indexed | 2024-03-06T13:12:40Z |
| publishDate | 2022 |
| record_format | dspace |
| spelling | UMPir398962024-01-08T07:35:37Z http://umpir.ump.edu.my/id/eprint/39896/ Analysis Of Personal Protective Equipment Classification Method Using Deep Learning Siti Zahrah Nur Ain, Silopung TA Engineering (General). Civil engineering (General) TK Electrical engineering. Electronics Nuclear engineering Personal Protective Equipment (PPE) is equipment worn in the workplace in order to reduce the exposure to hazards that causing serious injuries and illness to people. In this project, there are five types of PPE to be considered which is also recognized by Occupational Safety and Health Administration (OSHA) such as face mask, face shield, safety goggle, safety helmet and safety jacket. To avoid a tedious work in manually checking whether workers wear PPE or not, an automatic PPE classifier is constructed by utilizing a deep learning algorithm called Convolutional Neural Network (CNN). This classification is performed using Anaconda and Jupyter Notebok Software that use Python as the programming language. There are total of 7500 images in the PPE dataset, 6000 images for training with and another 1500 images for testing. The classification is done in two ways, one is by classifying the PPE into 5 classes and another one is by classifying into binary class after the best combined parameters are obtained using multiple training and testing by changing the parameters such as epoch, activation function, optimizer and filter layer. By using classifying into 5 classes, the final training accuracy is 89.39% and testing accuracy of 62.23%. On the other hand, by classifying the PPE into binary class, the PPE has final testing accuracy up 88% for all PPE. Face mask has final accuracy of 95.60%, face shield 94.32%, safety goggle 89.79%, safety helmet 98.90% and lastly safety jacket has 88.45% testing accuracy. Based on the result, CNN algorithm is a good algorithm as the binary classification of PPE achieved high accuracy result. 2022-06 Undergraduates Project Papers NonPeerReviewed pdf en http://umpir.ump.edu.my/id/eprint/39896/1/EA18087-SITI%20ZAHRAH%20NUR%20AIN-THESIS%20-%20Siti%20Zahrah.pdf Siti Zahrah Nur Ain, Silopung (2022) Analysis Of Personal Protective Equipment Classification Method Using Deep Learning. College of Engineering, Universiti Malaysia Pahang Al-Sultan Abdullah. |
| spellingShingle | TA Engineering (General). Civil engineering (General) TK Electrical engineering. Electronics Nuclear engineering Siti Zahrah Nur Ain, Silopung Analysis Of Personal Protective Equipment Classification Method Using Deep Learning |
| title | Analysis Of Personal Protective Equipment Classification Method Using Deep Learning |
| title_full | Analysis Of Personal Protective Equipment Classification Method Using Deep Learning |
| title_fullStr | Analysis Of Personal Protective Equipment Classification Method Using Deep Learning |
| title_full_unstemmed | Analysis Of Personal Protective Equipment Classification Method Using Deep Learning |
| title_short | Analysis Of Personal Protective Equipment Classification Method Using Deep Learning |
| title_sort | analysis of personal protective equipment classification method using deep learning |
| topic | TA Engineering (General). Civil engineering (General) TK Electrical engineering. Electronics Nuclear engineering |
| url | http://umpir.ump.edu.my/id/eprint/39896/1/EA18087-SITI%20ZAHRAH%20NUR%20AIN-THESIS%20-%20Siti%20Zahrah.pdf |
| work_keys_str_mv | AT sitizahrahnurainsilopung analysisofpersonalprotectiveequipmentclassificationmethodusingdeeplearning |