Vision Based Assistive System for Fall Detection
Shripad S. Bhatlawande, Dhawal Khapre, Mahesh Kinge, Tejas Khairnar · 2022 2nd International Conference on Intelligent Technologies (CONIT) · 2022
Falling is one of the most significant major accidental concerns for susceptible individuals like the elderly and vision challenged individuals. The field of computer vision has been garnering a lot of attention due to its enormous number of applications such as object detection and object recognition. Scientific experts have extensively researched falls as a result of technological improvements in order to reduce significant repercussions and bad effects. The project will assist visually challenged individuals if there would be a circumstance of fall. The identification of human falls is the subject of this research where SIFT has been used for feature extraction, K Means for feature selection and PCA for dimensionality reduction in this work. Classifiers such as KNN, Naive Bayes, Random Forest, SVM and Decision Tree have been implemented for prediction purpose. Recall, F1 score, accuracy, and precision are the performance indicators for the solution. Random Forest gave the highest accuracy on the dataset with an accuracy equal to 93.14%.