SlideShare a Scribd company logo
1 of 5
Base paper Title: A Novel Approach for Disaster Victim Detection Under Debris
Environments Using Decision Tree Algorithms With Deep Learning Features
Modified Title: A Novel Method Using Decision Tree Algorithms with Deep Learning
Features for Disaster Victim Detection in Debris Environments
Abstract
Search and Rescue operations for victim identification in an unstructured collapsed
building are high-risk and time-consuming. The possibility of saving a victim is high only
during the first 48 hours, and then the prospect tends to zero. The faster the response and
identification, the sooner the victim can be taken to medical assistance. Combining mobile
robots with practical Artificial Intelligence (AI) driven Human Victim Detection (HVD)
systems managed by professional teams can considerably reduce this problem. In this paper,
we have developed a Transfer Learning-based Deep Learning approach to identify human
victims under collapsed building environments by integrating machine learning classification
algorithms. A custom-made human victim dataset was created with five class labels: head,
hand, leg, upper body, and without the body. First, we extracted the class-wise features of the
dataset using fine-tuning-based transfer learning on ResNet-50 deep learning model. The
learned features of the model were then extracted, and then a feature selection was performed
using J48 to study the impact of feature reduction in classification. Several decision tree
algorithms, including decision stump, hoeffiding tree, J48, Linear Model Tree (LMT), Random
Forest, Random Tree, Representative (REP) Tree, J48 graft, and other famous algorithms like
LibSVM, Logistic regression, Multilayer perceptron, BayesNet, Naive Bayes are then used to
perform the classification. The classification accuracy of the abovementioned algorithms is
compared to recommend the optimal approach for real-time use. The random tree approach
outperformed all other tree-based algorithms with a maximum classification accuracy of
99.53% and a computation time of 0.02 seconds.
Existing System
Massive earthquakes, floods, plane crashes, tsunamis, and building collapses are only
a few examples of the many catastrophic natural and human-caused disasters that plague the
world. Disaster management is essential for reducing and avoiding the losses these catastrophes
cause. Fig.1 depicts the mortality statistics in building collapses caused by earthquakes and
other factors worldwide over the last few years. According to the emergency relief cycle [1],
the four stages of disaster management include prevention, preparedness, reaction, and
recovery. This work comes under the preparedness and reaction phases of the disaster
management cycle to aid fast rescue assistance for identifying victims. This work is a
continuation of our previous work [2] as part of our motto to develop a snake-like robot for
rescue assistance for human victim detection in earthquake environments. According to Urban
Search And Rescue (USAR), the chances of saving a victim are only good for the first 48 hours
of the rescue operation, after which the probability is almost zero. Hence, the faster the rescue
mission higher the probability of saving a life. Victim identification in the collapsed building
environment is one of the chief challenges for responders. One of the main concerns is the
accuracy of identification. Accurately identifying victims in this situation is often difficult
because of the uncertainties in the victims’ bodies and the surrounding environment. Despite
these potential limitations, this work tries to identify victims using the information available to
responders based on human physical anatomy. A detailed study of human identification
methods is discussed in the next section. RGB and thermal image-based datasets are more
useful in search and rescue scenarios involving the detection of human victims [3]. An RGB
image-based custom dataset is employed in this work for HVI in collapsed, unstructured
building scenarios. Once we have a dataset, its features must be learned to identify the victim
accurately. Deep learning (DL) has emerged as a powerful tool for computer vision
applications, including image classification and object detection. The most common deep
learning models are convolutional neural networks (CNNs), which can be applied to many data
types [4]. CNNs have become increasingly popular due to their ability to learn and extract
relevant features from raw image data and classify them into different categories based on those
features without manual feature engineering.
Drawback in Existing System
 Generalization Issues:
If the training data is not representative of the various scenarios that can occur in real
disaster situations, the model may struggle to generalize to new and unseen conditions.
This lack of generalization can limit the model's effectiveness in different disaster
environments.
 Computational Complexity:
Deep learning models, especially those with complex architectures, can be
computationally expensive and may require significant processing power.
Implementing such models in real-time disaster scenarios, where resources may be
limited, could be a practical challenge.
 Interpretability:
While decision trees are generally interpretable, the inclusion of deep learning
features might introduce complexity and reduce the interpretability of the model.
Understanding the decision-making process of the model is crucial, especially in critical
applications like disaster response.
 Integration with Existing Systems:
Integrating a novel approach into existing disaster response systems may pose
challenges. Compatibility, scalability, and communication with other components of
the disaster management infrastructure need to be considered.
Proposed System
 Data Acquisition:
Gather data from disaster scenarios involving debris environments. This dataset
should include information about the presence or absence of disaster victims,
environmental conditions, and types of debris.
 Evaluation Metrics:
Evaluate the model's performance using metrics such as precision, recall, F1 score,
and AUC-ROC on a separate test set. This step ensures that the model can generalize
well to new and unseen data.
 Interpretability Analysis:
Analyze the interpretability of the decision tree model, ensuring that the decision rules
can be understood by domain experts and stakeholders. The interpretability is crucial
for building trust in the system and facilitating collaboration with disaster response
teams.
 Integration with Disaster Response Infrastructure:
Ensure seamless integration with existing disaster response infrastructure,
communication systems, and protocols. The system should complement and enhance
the capabilities of human responders rather than replace them.
Algorithm
 Data Preprocessing:
Clean and preprocess the dataset to handle missing values, outliers, and other data
anomalies. This may involve normalizing numerical features, encoding categorical
variables, and addressing imbalances in the class distribution.
 Deep Learning Integration:
Incorporate deep learning features into the decision tree model. This can involve
using representations learned by a neural network as additional input features to
enhance the decision-making process of the tree.
 Evaluation:
Evaluate the performance of the model using a separate test set that was not used
during training. Common evaluation metrics may include precision, recall, F1 score,
and area under the receiver operating characteristic curve (AUC-ROC).
Advantages
 Feature Extraction with Deep Learning:
Integration of deep learning features allows the model to automatically learn and
extract relevant representations from the input data. This can be particularly beneficial
in capturing complex patterns and relationships within debris environments that may
be challenging for traditional handcrafted features.
 Improved Accuracy:
The integration of deep learning features has the potential to enhance the overall
accuracy of the model. Deep learning excels at capturing hierarchical and abstract
features, which can contribute to better discrimination between disaster victims and
non-victims in debris environments.
 Incorporation of Domain Knowledge:
Decision tree models allow for the incorporation of domain knowledge into the
decision-making process. Experts can contribute to the creation of decision rules and
interpretability, ensuring that the model aligns with practical insights from the field of
disaster response.
 Potential for Real-time Detection:
Decision trees, being relatively fast to evaluate, coupled with the potential for
optimized deep learning features, can enable real-time victim detection in disaster
scenarios. This rapid response is critical for timely and effective disaster management.
Software Specification
 Processor : I3 core processor
 Ram : 4 GB
 Hard disk : 500 GB
Software Specification
 Operating System : Windows 10 /11
 Frond End : Python
 Back End : Mysql Server
 IDE Tools : Pycharm

More Related Content

Similar to A Novel Approach for Disaster Victim Detection Under Debris Environments Using Decision Tree Algorithms With Deep Learning Features.docx

GeoAI: A Model-Agnostic Meta-Ensemble Zero-Shot Learning Method for Hyperspec...
GeoAI: A Model-Agnostic Meta-Ensemble Zero-Shot Learning Method for Hyperspec...GeoAI: A Model-Agnostic Meta-Ensemble Zero-Shot Learning Method for Hyperspec...
GeoAI: A Model-Agnostic Meta-Ensemble Zero-Shot Learning Method for Hyperspec...
Konstantinos Demertzis
 
Dynamic Rule Base Construction and Maintenance Scheme for Disease Prediction
Dynamic Rule Base Construction and Maintenance Scheme for Disease PredictionDynamic Rule Base Construction and Maintenance Scheme for Disease Prediction
Dynamic Rule Base Construction and Maintenance Scheme for Disease Prediction
ijsrd.com
 
ADDRESSING IMBALANCED CLASSES PROBLEM OF INTRUSION DETECTION SYSTEM USING WEI...
ADDRESSING IMBALANCED CLASSES PROBLEM OF INTRUSION DETECTION SYSTEM USING WEI...ADDRESSING IMBALANCED CLASSES PROBLEM OF INTRUSION DETECTION SYSTEM USING WEI...
ADDRESSING IMBALANCED CLASSES PROBLEM OF INTRUSION DETECTION SYSTEM USING WEI...
IJCNCJournal
 
A Review on Reasoning System, Types, and Tools and Need for Hybrid Reasoning
A Review on Reasoning System, Types, and Tools and Need for Hybrid ReasoningA Review on Reasoning System, Types, and Tools and Need for Hybrid Reasoning
A Review on Reasoning System, Types, and Tools and Need for Hybrid Reasoning
BRNSSPublicationHubI
 

Similar to A Novel Approach for Disaster Victim Detection Under Debris Environments Using Decision Tree Algorithms With Deep Learning Features.docx (20)

Comparative Analysis of Computational Intelligence Paradigms in WSN: Review
Comparative Analysis of Computational Intelligence Paradigms in WSN: ReviewComparative Analysis of Computational Intelligence Paradigms in WSN: Review
Comparative Analysis of Computational Intelligence Paradigms in WSN: Review
 
Pattern recognition using context dependent memory model (cdmm) in multimodal...
Pattern recognition using context dependent memory model (cdmm) in multimodal...Pattern recognition using context dependent memory model (cdmm) in multimodal...
Pattern recognition using context dependent memory model (cdmm) in multimodal...
 
Dynamic responsibilities assignment in critical electronic institutions
Dynamic responsibilities assignment in critical electronic institutionsDynamic responsibilities assignment in critical electronic institutions
Dynamic responsibilities assignment in critical electronic institutions
 
GeoAI: A Model-Agnostic Meta-Ensemble Zero-Shot Learning Method for Hyperspec...
GeoAI: A Model-Agnostic Meta-Ensemble Zero-Shot Learning Method for Hyperspec...GeoAI: A Model-Agnostic Meta-Ensemble Zero-Shot Learning Method for Hyperspec...
GeoAI: A Model-Agnostic Meta-Ensemble Zero-Shot Learning Method for Hyperspec...
 
IRJET- Comparative Study of Different Techniques for Text as Well as Object D...
IRJET- Comparative Study of Different Techniques for Text as Well as Object D...IRJET- Comparative Study of Different Techniques for Text as Well as Object D...
IRJET- Comparative Study of Different Techniques for Text as Well as Object D...
 
Fake News Detection using Deep Learning
Fake News Detection using Deep LearningFake News Detection using Deep Learning
Fake News Detection using Deep Learning
 
Exploration exploitation trade off in mobile context-aware recommender systems
Exploration  exploitation trade off in mobile context-aware recommender systemsExploration  exploitation trade off in mobile context-aware recommender systems
Exploration exploitation trade off in mobile context-aware recommender systems
 
A Survey of Security of Multimodal Biometric Systems
A Survey of Security of Multimodal Biometric SystemsA Survey of Security of Multimodal Biometric Systems
A Survey of Security of Multimodal Biometric Systems
 
Machine Learning in Malware Detection
Machine Learning in Malware DetectionMachine Learning in Malware Detection
Machine Learning in Malware Detection
 
A USER PROFILE BASED ACCESS CONTROL MODEL AND ARCHITECTURE
A USER PROFILE BASED ACCESS CONTROL MODEL  AND ARCHITECTUREA USER PROFILE BASED ACCESS CONTROL MODEL  AND ARCHITECTURE
A USER PROFILE BASED ACCESS CONTROL MODEL AND ARCHITECTURE
 
Detecting anomalies in security cameras with 3D-convolutional neural network ...
Detecting anomalies in security cameras with 3D-convolutional neural network ...Detecting anomalies in security cameras with 3D-convolutional neural network ...
Detecting anomalies in security cameras with 3D-convolutional neural network ...
 
INTELLIGENT MALWARE DETECTION USING EXTREME LEARNING MACHINE
INTELLIGENT MALWARE DETECTION USING EXTREME LEARNING MACHINEINTELLIGENT MALWARE DETECTION USING EXTREME LEARNING MACHINE
INTELLIGENT MALWARE DETECTION USING EXTREME LEARNING MACHINE
 
Facial Expression Identification System
Facial Expression Identification SystemFacial Expression Identification System
Facial Expression Identification System
 
A Development Shell For Cooperative Problem-Solving Environments
A Development Shell For Cooperative Problem-Solving EnvironmentsA Development Shell For Cooperative Problem-Solving Environments
A Development Shell For Cooperative Problem-Solving Environments
 
IRJET- Analysing Wound Area Measurement using Android App
IRJET- Analysing Wound Area Measurement using Android AppIRJET- Analysing Wound Area Measurement using Android App
IRJET- Analysing Wound Area Measurement using Android App
 
Dynamic Rule Base Construction and Maintenance Scheme for Disease Prediction
Dynamic Rule Base Construction and Maintenance Scheme for Disease PredictionDynamic Rule Base Construction and Maintenance Scheme for Disease Prediction
Dynamic Rule Base Construction and Maintenance Scheme for Disease Prediction
 
ADDRESSING IMBALANCED CLASSES PROBLEM OF INTRUSION DETECTION SYSTEM USING WEI...
ADDRESSING IMBALANCED CLASSES PROBLEM OF INTRUSION DETECTION SYSTEM USING WEI...ADDRESSING IMBALANCED CLASSES PROBLEM OF INTRUSION DETECTION SYSTEM USING WEI...
ADDRESSING IMBALANCED CLASSES PROBLEM OF INTRUSION DETECTION SYSTEM USING WEI...
 
Intelligent interaction for smart work [
Intelligent interaction for smart work [Intelligent interaction for smart work [
Intelligent interaction for smart work [
 
Image Forgery Detection Methods- A Review
Image Forgery Detection Methods- A ReviewImage Forgery Detection Methods- A Review
Image Forgery Detection Methods- A Review
 
A Review on Reasoning System, Types, and Tools and Need for Hybrid Reasoning
A Review on Reasoning System, Types, and Tools and Need for Hybrid ReasoningA Review on Reasoning System, Types, and Tools and Need for Hybrid Reasoning
A Review on Reasoning System, Types, and Tools and Need for Hybrid Reasoning
 

More from Shakas Technologies

More from Shakas Technologies (20)

A Review on Deep-Learning-Based Cyberbullying Detection
A Review on Deep-Learning-Based Cyberbullying DetectionA Review on Deep-Learning-Based Cyberbullying Detection
A Review on Deep-Learning-Based Cyberbullying Detection
 
A Personal Privacy Data Protection Scheme for Encryption and Revocation of Hi...
A Personal Privacy Data Protection Scheme for Encryption and Revocation of Hi...A Personal Privacy Data Protection Scheme for Encryption and Revocation of Hi...
A Personal Privacy Data Protection Scheme for Encryption and Revocation of Hi...
 
A Novel Framework for Credit Card.
A Novel Framework for Credit Card.A Novel Framework for Credit Card.
A Novel Framework for Credit Card.
 
A Comparative Analysis of Sampling Techniques for Click-Through Rate Predicti...
A Comparative Analysis of Sampling Techniques for Click-Through Rate Predicti...A Comparative Analysis of Sampling Techniques for Click-Through Rate Predicti...
A Comparative Analysis of Sampling Techniques for Click-Through Rate Predicti...
 
NS2 Final Year Project Titles 2023- 2024
NS2 Final Year Project Titles 2023- 2024NS2 Final Year Project Titles 2023- 2024
NS2 Final Year Project Titles 2023- 2024
 
MATLAB Final Year IEEE Project Titles 2023-2024
MATLAB Final Year IEEE Project Titles 2023-2024MATLAB Final Year IEEE Project Titles 2023-2024
MATLAB Final Year IEEE Project Titles 2023-2024
 
Latest Python IEEE Project Titles 2023-2024
Latest Python IEEE Project Titles 2023-2024Latest Python IEEE Project Titles 2023-2024
Latest Python IEEE Project Titles 2023-2024
 
EMOTION RECOGNITION BY TEXTUAL TWEETS CLASSIFICATION USING VOTING CLASSIFIER ...
EMOTION RECOGNITION BY TEXTUAL TWEETS CLASSIFICATION USING VOTING CLASSIFIER ...EMOTION RECOGNITION BY TEXTUAL TWEETS CLASSIFICATION USING VOTING CLASSIFIER ...
EMOTION RECOGNITION BY TEXTUAL TWEETS CLASSIFICATION USING VOTING CLASSIFIER ...
 
CYBER THREAT INTELLIGENCE MINING FOR PROACTIVE CYBERSECURITY DEFENSE
CYBER THREAT INTELLIGENCE MINING FOR PROACTIVE CYBERSECURITY DEFENSECYBER THREAT INTELLIGENCE MINING FOR PROACTIVE CYBERSECURITY DEFENSE
CYBER THREAT INTELLIGENCE MINING FOR PROACTIVE CYBERSECURITY DEFENSE
 
Detecting Mental Disorders in social Media through Emotional patterns-The cas...
Detecting Mental Disorders in social Media through Emotional patterns-The cas...Detecting Mental Disorders in social Media through Emotional patterns-The cas...
Detecting Mental Disorders in social Media through Emotional patterns-The cas...
 
COMMERCE FAKE PRODUCT REVIEWS MONITORING AND DETECTION
COMMERCE FAKE PRODUCT REVIEWS MONITORING AND DETECTIONCOMMERCE FAKE PRODUCT REVIEWS MONITORING AND DETECTION
COMMERCE FAKE PRODUCT REVIEWS MONITORING AND DETECTION
 
CO2 EMISSION RATING BY VEHICLES USING DATA SCIENCE
CO2 EMISSION RATING BY VEHICLES USING DATA SCIENCECO2 EMISSION RATING BY VEHICLES USING DATA SCIENCE
CO2 EMISSION RATING BY VEHICLES USING DATA SCIENCE
 
Toward Effective Evaluation of Cyber Defense Threat Based Adversary Emulation...
Toward Effective Evaluation of Cyber Defense Threat Based Adversary Emulation...Toward Effective Evaluation of Cyber Defense Threat Based Adversary Emulation...
Toward Effective Evaluation of Cyber Defense Threat Based Adversary Emulation...
 
Optimizing Numerical Weather Prediction Model Performance Using Machine Learn...
Optimizing Numerical Weather Prediction Model Performance Using Machine Learn...Optimizing Numerical Weather Prediction Model Performance Using Machine Learn...
Optimizing Numerical Weather Prediction Model Performance Using Machine Learn...
 
Nature-Based Prediction Model of Bug Reports Based on Ensemble Machine Learni...
Nature-Based Prediction Model of Bug Reports Based on Ensemble Machine Learni...Nature-Based Prediction Model of Bug Reports Based on Ensemble Machine Learni...
Nature-Based Prediction Model of Bug Reports Based on Ensemble Machine Learni...
 
Multi-Class Stress Detection Through Heart Rate Variability A Deep Neural Net...
Multi-Class Stress Detection Through Heart Rate Variability A Deep Neural Net...Multi-Class Stress Detection Through Heart Rate Variability A Deep Neural Net...
Multi-Class Stress Detection Through Heart Rate Variability A Deep Neural Net...
 
Identifying Hot Topic Trends in Streaming Text Data Using News Sequential Evo...
Identifying Hot Topic Trends in Streaming Text Data Using News Sequential Evo...Identifying Hot Topic Trends in Streaming Text Data Using News Sequential Evo...
Identifying Hot Topic Trends in Streaming Text Data Using News Sequential Evo...
 
Fighting Money Laundering With Statistics and Machine Learning.docx
Fighting Money Laundering With Statistics and Machine Learning.docxFighting Money Laundering With Statistics and Machine Learning.docx
Fighting Money Laundering With Statistics and Machine Learning.docx
 
Explainable Artificial Intelligence for Patient Safety A Review of Applicatio...
Explainable Artificial Intelligence for Patient Safety A Review of Applicatio...Explainable Artificial Intelligence for Patient Safety A Review of Applicatio...
Explainable Artificial Intelligence for Patient Safety A Review of Applicatio...
 
Ensemble Deep Learning-Based Prediction of Fraudulent Cryptocurrency Transact...
Ensemble Deep Learning-Based Prediction of Fraudulent Cryptocurrency Transact...Ensemble Deep Learning-Based Prediction of Fraudulent Cryptocurrency Transact...
Ensemble Deep Learning-Based Prediction of Fraudulent Cryptocurrency Transact...
 

Recently uploaded

Orientation Canvas Course Presentation.pdf
Orientation Canvas Course Presentation.pdfOrientation Canvas Course Presentation.pdf
Orientation Canvas Course Presentation.pdf
Elizabeth Walsh
 
MuleSoft Integration with AWS Textract | Calling AWS Textract API |AWS - Clou...
MuleSoft Integration with AWS Textract | Calling AWS Textract API |AWS - Clou...MuleSoft Integration with AWS Textract | Calling AWS Textract API |AWS - Clou...
MuleSoft Integration with AWS Textract | Calling AWS Textract API |AWS - Clou...
MysoreMuleSoftMeetup
 

Recently uploaded (20)

Details on CBSE Compartment Exam.pptx1111
Details on CBSE Compartment Exam.pptx1111Details on CBSE Compartment Exam.pptx1111
Details on CBSE Compartment Exam.pptx1111
 
Tatlong Kwento ni Lola basyang-1.pdf arts
Tatlong Kwento ni Lola basyang-1.pdf artsTatlong Kwento ni Lola basyang-1.pdf arts
Tatlong Kwento ni Lola basyang-1.pdf arts
 
UGC NET Paper 1 Unit 7 DATA INTERPRETATION.pdf
UGC NET Paper 1 Unit 7 DATA INTERPRETATION.pdfUGC NET Paper 1 Unit 7 DATA INTERPRETATION.pdf
UGC NET Paper 1 Unit 7 DATA INTERPRETATION.pdf
 
Towards a code of practice for AI in AT.pptx
Towards a code of practice for AI in AT.pptxTowards a code of practice for AI in AT.pptx
Towards a code of practice for AI in AT.pptx
 
Simple, Complex, and Compound Sentences Exercises.pdf
Simple, Complex, and Compound Sentences Exercises.pdfSimple, Complex, and Compound Sentences Exercises.pdf
Simple, Complex, and Compound Sentences Exercises.pdf
 
How to Add a Tool Tip to a Field in Odoo 17
How to Add a Tool Tip to a Field in Odoo 17How to Add a Tool Tip to a Field in Odoo 17
How to Add a Tool Tip to a Field in Odoo 17
 
Ernest Hemingway's For Whom the Bell Tolls
Ernest Hemingway's For Whom the Bell TollsErnest Hemingway's For Whom the Bell Tolls
Ernest Hemingway's For Whom the Bell Tolls
 
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptxHMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
HMCS Max Bernays Pre-Deployment Brief (May 2024).pptx
 
NO1 Top Black Magic Specialist In Lahore Black magic In Pakistan Kala Ilam Ex...
NO1 Top Black Magic Specialist In Lahore Black magic In Pakistan Kala Ilam Ex...NO1 Top Black Magic Specialist In Lahore Black magic In Pakistan Kala Ilam Ex...
NO1 Top Black Magic Specialist In Lahore Black magic In Pakistan Kala Ilam Ex...
 
TỔNG HỢP HƠN 100 ĐỀ THI THỬ TỐT NGHIỆP THPT TOÁN 2024 - TỪ CÁC TRƯỜNG, TRƯỜNG...
TỔNG HỢP HƠN 100 ĐỀ THI THỬ TỐT NGHIỆP THPT TOÁN 2024 - TỪ CÁC TRƯỜNG, TRƯỜNG...TỔNG HỢP HƠN 100 ĐỀ THI THỬ TỐT NGHIỆP THPT TOÁN 2024 - TỪ CÁC TRƯỜNG, TRƯỜNG...
TỔNG HỢP HƠN 100 ĐỀ THI THỬ TỐT NGHIỆP THPT TOÁN 2024 - TỪ CÁC TRƯỜNG, TRƯỜNG...
 
PANDITA RAMABAI- Indian political thought GENDER.pptx
PANDITA RAMABAI- Indian political thought GENDER.pptxPANDITA RAMABAI- Indian political thought GENDER.pptx
PANDITA RAMABAI- Indian political thought GENDER.pptx
 
When Quality Assurance Meets Innovation in Higher Education - Report launch w...
When Quality Assurance Meets Innovation in Higher Education - Report launch w...When Quality Assurance Meets Innovation in Higher Education - Report launch w...
When Quality Assurance Meets Innovation in Higher Education - Report launch w...
 
diagnosting testing bsc 2nd sem.pptx....
diagnosting testing bsc 2nd sem.pptx....diagnosting testing bsc 2nd sem.pptx....
diagnosting testing bsc 2nd sem.pptx....
 
Orientation Canvas Course Presentation.pdf
Orientation Canvas Course Presentation.pdfOrientation Canvas Course Presentation.pdf
Orientation Canvas Course Presentation.pdf
 
MuleSoft Integration with AWS Textract | Calling AWS Textract API |AWS - Clou...
MuleSoft Integration with AWS Textract | Calling AWS Textract API |AWS - Clou...MuleSoft Integration with AWS Textract | Calling AWS Textract API |AWS - Clou...
MuleSoft Integration with AWS Textract | Calling AWS Textract API |AWS - Clou...
 
AIM of Education-Teachers Training-2024.ppt
AIM of Education-Teachers Training-2024.pptAIM of Education-Teachers Training-2024.ppt
AIM of Education-Teachers Training-2024.ppt
 
How to Manage Call for Tendor in Odoo 17
How to Manage Call for Tendor in Odoo 17How to Manage Call for Tendor in Odoo 17
How to Manage Call for Tendor in Odoo 17
 
80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...
80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...
80 ĐỀ THI THỬ TUYỂN SINH TIẾNG ANH VÀO 10 SỞ GD – ĐT THÀNH PHỐ HỒ CHÍ MINH NĂ...
 
Diuretic, Hypoglycemic and Limit test of Heavy metals and Arsenic.-1.pdf
Diuretic, Hypoglycemic and Limit test of Heavy metals and Arsenic.-1.pdfDiuretic, Hypoglycemic and Limit test of Heavy metals and Arsenic.-1.pdf
Diuretic, Hypoglycemic and Limit test of Heavy metals and Arsenic.-1.pdf
 
OSCM Unit 2_Operations Processes & Systems
OSCM Unit 2_Operations Processes & SystemsOSCM Unit 2_Operations Processes & Systems
OSCM Unit 2_Operations Processes & Systems
 

A Novel Approach for Disaster Victim Detection Under Debris Environments Using Decision Tree Algorithms With Deep Learning Features.docx

  • 1. Base paper Title: A Novel Approach for Disaster Victim Detection Under Debris Environments Using Decision Tree Algorithms With Deep Learning Features Modified Title: A Novel Method Using Decision Tree Algorithms with Deep Learning Features for Disaster Victim Detection in Debris Environments Abstract Search and Rescue operations for victim identification in an unstructured collapsed building are high-risk and time-consuming. The possibility of saving a victim is high only during the first 48 hours, and then the prospect tends to zero. The faster the response and identification, the sooner the victim can be taken to medical assistance. Combining mobile robots with practical Artificial Intelligence (AI) driven Human Victim Detection (HVD) systems managed by professional teams can considerably reduce this problem. In this paper, we have developed a Transfer Learning-based Deep Learning approach to identify human victims under collapsed building environments by integrating machine learning classification algorithms. A custom-made human victim dataset was created with five class labels: head, hand, leg, upper body, and without the body. First, we extracted the class-wise features of the dataset using fine-tuning-based transfer learning on ResNet-50 deep learning model. The learned features of the model were then extracted, and then a feature selection was performed using J48 to study the impact of feature reduction in classification. Several decision tree algorithms, including decision stump, hoeffiding tree, J48, Linear Model Tree (LMT), Random Forest, Random Tree, Representative (REP) Tree, J48 graft, and other famous algorithms like LibSVM, Logistic regression, Multilayer perceptron, BayesNet, Naive Bayes are then used to perform the classification. The classification accuracy of the abovementioned algorithms is compared to recommend the optimal approach for real-time use. The random tree approach outperformed all other tree-based algorithms with a maximum classification accuracy of 99.53% and a computation time of 0.02 seconds. Existing System Massive earthquakes, floods, plane crashes, tsunamis, and building collapses are only a few examples of the many catastrophic natural and human-caused disasters that plague the world. Disaster management is essential for reducing and avoiding the losses these catastrophes cause. Fig.1 depicts the mortality statistics in building collapses caused by earthquakes and
  • 2. other factors worldwide over the last few years. According to the emergency relief cycle [1], the four stages of disaster management include prevention, preparedness, reaction, and recovery. This work comes under the preparedness and reaction phases of the disaster management cycle to aid fast rescue assistance for identifying victims. This work is a continuation of our previous work [2] as part of our motto to develop a snake-like robot for rescue assistance for human victim detection in earthquake environments. According to Urban Search And Rescue (USAR), the chances of saving a victim are only good for the first 48 hours of the rescue operation, after which the probability is almost zero. Hence, the faster the rescue mission higher the probability of saving a life. Victim identification in the collapsed building environment is one of the chief challenges for responders. One of the main concerns is the accuracy of identification. Accurately identifying victims in this situation is often difficult because of the uncertainties in the victims’ bodies and the surrounding environment. Despite these potential limitations, this work tries to identify victims using the information available to responders based on human physical anatomy. A detailed study of human identification methods is discussed in the next section. RGB and thermal image-based datasets are more useful in search and rescue scenarios involving the detection of human victims [3]. An RGB image-based custom dataset is employed in this work for HVI in collapsed, unstructured building scenarios. Once we have a dataset, its features must be learned to identify the victim accurately. Deep learning (DL) has emerged as a powerful tool for computer vision applications, including image classification and object detection. The most common deep learning models are convolutional neural networks (CNNs), which can be applied to many data types [4]. CNNs have become increasingly popular due to their ability to learn and extract relevant features from raw image data and classify them into different categories based on those features without manual feature engineering. Drawback in Existing System  Generalization Issues: If the training data is not representative of the various scenarios that can occur in real disaster situations, the model may struggle to generalize to new and unseen conditions. This lack of generalization can limit the model's effectiveness in different disaster environments.
  • 3.  Computational Complexity: Deep learning models, especially those with complex architectures, can be computationally expensive and may require significant processing power. Implementing such models in real-time disaster scenarios, where resources may be limited, could be a practical challenge.  Interpretability: While decision trees are generally interpretable, the inclusion of deep learning features might introduce complexity and reduce the interpretability of the model. Understanding the decision-making process of the model is crucial, especially in critical applications like disaster response.  Integration with Existing Systems: Integrating a novel approach into existing disaster response systems may pose challenges. Compatibility, scalability, and communication with other components of the disaster management infrastructure need to be considered. Proposed System  Data Acquisition: Gather data from disaster scenarios involving debris environments. This dataset should include information about the presence or absence of disaster victims, environmental conditions, and types of debris.  Evaluation Metrics: Evaluate the model's performance using metrics such as precision, recall, F1 score, and AUC-ROC on a separate test set. This step ensures that the model can generalize well to new and unseen data.  Interpretability Analysis: Analyze the interpretability of the decision tree model, ensuring that the decision rules can be understood by domain experts and stakeholders. The interpretability is crucial
  • 4. for building trust in the system and facilitating collaboration with disaster response teams.  Integration with Disaster Response Infrastructure: Ensure seamless integration with existing disaster response infrastructure, communication systems, and protocols. The system should complement and enhance the capabilities of human responders rather than replace them. Algorithm  Data Preprocessing: Clean and preprocess the dataset to handle missing values, outliers, and other data anomalies. This may involve normalizing numerical features, encoding categorical variables, and addressing imbalances in the class distribution.  Deep Learning Integration: Incorporate deep learning features into the decision tree model. This can involve using representations learned by a neural network as additional input features to enhance the decision-making process of the tree.  Evaluation: Evaluate the performance of the model using a separate test set that was not used during training. Common evaluation metrics may include precision, recall, F1 score, and area under the receiver operating characteristic curve (AUC-ROC). Advantages  Feature Extraction with Deep Learning: Integration of deep learning features allows the model to automatically learn and extract relevant representations from the input data. This can be particularly beneficial in capturing complex patterns and relationships within debris environments that may be challenging for traditional handcrafted features.
  • 5.  Improved Accuracy: The integration of deep learning features has the potential to enhance the overall accuracy of the model. Deep learning excels at capturing hierarchical and abstract features, which can contribute to better discrimination between disaster victims and non-victims in debris environments.  Incorporation of Domain Knowledge: Decision tree models allow for the incorporation of domain knowledge into the decision-making process. Experts can contribute to the creation of decision rules and interpretability, ensuring that the model aligns with practical insights from the field of disaster response.  Potential for Real-time Detection: Decision trees, being relatively fast to evaluate, coupled with the potential for optimized deep learning features, can enable real-time victim detection in disaster scenarios. This rapid response is critical for timely and effective disaster management. Software Specification  Processor : I3 core processor  Ram : 4 GB  Hard disk : 500 GB Software Specification  Operating System : Windows 10 /11  Frond End : Python  Back End : Mysql Server  IDE Tools : Pycharm