Leveraging retinanet based object detection model for assisting visually impaired individuals with metaheuristic optimization algorithm DOI Creative Commons
Alaa O. Khadidos, Ayman Yafoz

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: May 8, 2025

Visually impaired individuals suffer many problems in handling their everyday activities like road crossing, writing, finding an object, reading, and so on. However, navigation methods are available, efficient object detection (OD) for visually people (VIP) need to be improved. OD is the most crucial contribution of computer vision (CV) plays important part recognizing objects image. The aged challenged can identify different accurately under some constraint features scaled, occlusion, illuminated, blurred nature real world. A considerable number studies performed domain real-time recognition using deep learning (DL). DL-based remove autonomously classify detect objects. This paper proposes a novel Object Detection Model Impaired Individuals with Metaheuristic Optimization Algorithm (ODMVII-MOA) technique. proposed ODMVII-MOA technique aims improve advanced techniques recognize disabled people. At first, image pre-processing stage applies Weiner filter (WF) method enhance quality by eliminating unwanted noise from data. Furthermore, RetinaNet utilized process locate within Besides, employs EfficientNetB0 feature extraction process. For classification process, LSTM-AE employed. Finally, Dandelion Optimizer (DO) adjusts hyperparameter range optimally results better performance classification. experimental validation model verified indoor dataset outcomes determined regarding measures. comparison study portrayed superior accuracy value 99.69% over existing techniques.

Language: Английский

Leveraging retinanet based object detection model for assisting visually impaired individuals with metaheuristic optimization algorithm DOI Creative Commons
Alaa O. Khadidos, Ayman Yafoz

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: May 8, 2025

Visually impaired individuals suffer many problems in handling their everyday activities like road crossing, writing, finding an object, reading, and so on. However, navigation methods are available, efficient object detection (OD) for visually people (VIP) need to be improved. OD is the most crucial contribution of computer vision (CV) plays important part recognizing objects image. The aged challenged can identify different accurately under some constraint features scaled, occlusion, illuminated, blurred nature real world. A considerable number studies performed domain real-time recognition using deep learning (DL). DL-based remove autonomously classify detect objects. This paper proposes a novel Object Detection Model Impaired Individuals with Metaheuristic Optimization Algorithm (ODMVII-MOA) technique. proposed ODMVII-MOA technique aims improve advanced techniques recognize disabled people. At first, image pre-processing stage applies Weiner filter (WF) method enhance quality by eliminating unwanted noise from data. Furthermore, RetinaNet utilized process locate within Besides, employs EfficientNetB0 feature extraction process. For classification process, LSTM-AE employed. Finally, Dandelion Optimizer (DO) adjusts hyperparameter range optimally results better performance classification. experimental validation model verified indoor dataset outcomes determined regarding measures. comparison study portrayed superior accuracy value 99.69% over existing techniques.

Language: Английский

Citations

0