1. An intelligent framework to detect and generate alert while cattle lying on road in dangerous states using surveillance videos.
- Author
-
Kahlon, Gursimran Singh, Singh, Harnoor, Saini, Munish, and Kaur, Sandeep
- Subjects
VIDEO surveillance ,MASS surveillance ,CATTLE ,TRAFFIC accidents ,FEATURE extraction - Abstract
This paper presents the framework for early detection of cattle lying dead, blocking the road, or in a condition to cause an accident on highways based on video scrutiny. The suggested framework works with feature extraction, feature expression, and assessment criteria. It also includes a method that can identify the status of cattle on road i.e., cattle lying dead after an accident, a method for assessing stray cattle's blocking the road, and rules for assessing other dangerous states. The framework is trained over 6000 images covering almost all the highway collision possibilities and tested on real images depicting different scenarios on highways. The proposed framework can achieve the mAP of up to 95.10% with precision and recall values of 0.98 and 0.94 respectively, which outperforms all other approaches taken into consideration. The outcomes of the study will assist the concerned authorities to take preventive steps to avoid road accidents or traffic-related issues. [ABSTRACT FROM AUTHOR]
- Published
- 2023
- Full Text
- View/download PDF