Abstract
Elephant Entry in human settlements causes a major threat to Human–Elephant Conflict (HEC). As a result, humans and elephants both face unique problems, elephants destroy crops, houses, and many other things, while for protection of crops and villages humans use electrical fencing, throw stones, and use smoke and fire. Therefore, there needs to be research carried out to address these challenges while implementing autonomous elephant detection. (Objectives): To identify suitable and highly efficient Convolutional Neural Network (CNN) models for real-time elephant detection, it is a more accurate image processing algorithm that takes help of deep learning to carry out generative as well as descriptive jobs, by implementing various models of CNN and comparing the results among one other and displaying the results. (Methods): Literature review has been performed to identify suitable elephant detection models for real-time objects and different types of methods for detecting and tracking elephants. Following this, experiments have been conducted to evaluate the performance of the selected object detection models. Two methods are used in this experiment YOLO (You Look Only Once) algorithm that detect and identify things in picture and another one is SSD_efficientdets which is single shot detector but here the backbone networks are Efficientnets trained on ImageNet and the proposed BiFPN (Bi-directional feature Pyramid Network) network serves as a feature network. This BiFPN network intakes 3 to 7 features from the backbone network and applies bottom-up and top-down bi-directional feature fusion repeatedly (Results): You Look Only Once (YOLOv3) and SSD_efficientdet_d0_512x512 have been identified from the literature review as the most suitable and efficient algorithm for detecting elephants in real time. The detection performance of these algorithms has been calculated and compared with each other. The results have been presented. (Conclusion): The accuracy of YOLOv3 has been found to be better than the SSD_efficientdets_d0_512x512 model. Therefore, it has been concluded that YOLOv3 is the best model in the real-time detection of elephants. The results of YOLOv3 are better for classification performance compared to SSD_efficientdets_d0_512x512.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
Shukla, P, Dua, I., Raman, B., Mittal, A.: A computer vision framework for detecting and preventing human elephant collisions (2017). https://openaccess.thecvf.com/content_ICCV_2017_workshops/papers/w41/Shukla_A_Computer_Vision_ICCV_2017_paper.pdf
Sugumar, S.J., Jayaparvathy, R.: An early warning system for elephant intrusion along the forest border area. Curr. Sci. 5(1) (2018). ISSN No 2349-2503
Sheridan, N.K.: Elephant AI (2022). https://github.com/nksheridan/elephantAI
Bjorck, J., Rappazzo, B.H., Chen, D., Bernstein, R., Wrege, P., Gomes, C.: Automatic detection and compression for passive acoustic monitoring of the African forest elephant. In: Proceedings of the AAAI conference on artificial intelligence, vol. 33, pp. 476–484 (2019). https://doi.org/10.1609/aaai.v33i01.3301476
https://imadelhanafi.com/posts/object_detection_yolo_efficientdet_mobilenet/
Future Work
Future work can be done by testing models on videos and on live camera which models show best accuracy and to implement in real world in future we will try to implement code in NVIDIA Jetson Nano or on Raspberry pi.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2022 The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Patel, D., Sharma, S. (2022). Automated Detection of Elephant Using AI Techniques. In: Jabeen, S.D., Ali, J., Castillo, O. (eds) Soft Computing and Optimization. SCOTA 2021. Springer Proceedings in Mathematics & Statistics, vol 404. Springer, Singapore. https://doi.org/10.1007/978-981-19-6406-0_4
Download citation
DOI: https://doi.org/10.1007/978-981-19-6406-0_4
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-19-6405-3
Online ISBN: 978-981-19-6406-0
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)