Computer Vision based Animal Collision Avoidance Framework for Autonomous Vehicles

Gupta, Savyasachi; Chand, Dhananjai; Kavati, Ilaiah

doi:10.48550/arxiv.2012.10878

Cited by 2 publications

(2 citation statements)

References 9 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The authors achieved an accuracy of 94%. Gupta et al [44] used the Mask R-CNN model with a pre-trained network, ResNet-101, to detect two animal species (cows and dogs). They achieved an average precision of 79.47% and 81.09%, for detecting cows and dogs, respectively.…”

Section: B Animal Species Detectionmentioning

confidence: 99%

An Accurate and Fast Animal Species Detection System for Embedded Devices

Ibraheam

Gebali

2023

IEEE Access

View full text Add to dashboard Cite

Encounters between humans and wildlife often lead to injuries, especially in remote wilderness regions, and highways. Therefore, animal detection is a vital safety and wildlife conservation component that can mitigate the negative impacts of these encounters. Deep learning techniques have achieved the best results compared to other object detection techniques; however, they require many computations and parameters. A lightweight animal species detection model based on YOLOv2 was proposed. It was designed as a proof of concept of and as a first step to build a real-time mitigation system with embedded devices. Multi-level features merging is employed by adding a new pass-through layer to improve the feature extraction ability and accuracy of YOLOv2. Moreover, the two repeated 3 × 3 convolutional layers in the seventh block of the YOLOv2 architecture are removed to reduce computational complexity, and thus increase detection speed without reducing accuracy. Animal species detection methods based on regular Convolutional Neural Networks (CNNs) have been widely applied; however, these methods are difficult to adapt to geometric variations of animals in images. Thus, a modified YOLOv2 with the addition of deformable convolutional layers (DCLs) was proposed to resolve this issue. Our experimental results show that the proposed model outperforms the original YOLOv2 by 5.0% in accuracy and 12.0% in speed. Furthermore, our analysis shows that the modified YOLOv2 model is more suitable for deployment than YOLOv3 and YOLOv4 on embedded devices.

show abstract

Section: B Animal Species Detectionmentioning

confidence: 99%

An Accurate and Fast Animal Species Detection System for Embedded Devices

Ibraheam

Gebali

2023

IEEE Access

View full text Add to dashboard Cite

show abstract

“…Hough transform is another popular method for lane detection, which identifies lines in an image by converting the image space into a parameter space [14]. Although it is effective in detecting straight lines, the Hough transform struggles with curved lanes and requires additional preprocessing steps to address issues such as perspective distortion.…”

Section: A Traditional Approachesmentioning

confidence: 99%

Mask R-CNN Approach to Real-Time Lane Detection for Autonomous Vehicles

Abdrakhmanov,

Elemesova,

Zhussipbek

et al. 2023

IJACSA

View full text Add to dashboard Cite

The accurate and real-time detection of road lanes is crucial for the safe navigation of autonomous vehicles (AVs). This paper presents a novel approach to lane detection by leveraging the capabilities of the Mask Region-based Convolutional Neural Network (Mask R-CNN) model. Our method adapts Mask R-CNN to specifically address the challenges posed by diverse traffic scenarios and varying environmental conditions. We introduce a robust, efficient, and scalable architecture for lane detection, which segments the lane markings and generates precise boundaries for AVs to follow. We augment the model with a custom dataset, consisting of images collected from different geographical locations, weather conditions, and road types. This comprehensive dataset ensures the model's generalizability and adaptability to real-world conditions. We also introduce a multi-scale feature extraction technique, which improves the model's ability to detect lanes in both near and far fields of view. Our proposed method significantly outperforms existing state-of-the-art techniques in terms of accuracy, processing speed, and adaptability. Extensive experiments were conducted on public datasets and our custom dataset to validate the performance of the proposed method. Results demonstrate that our Mask R-CNN-based approach achieves high precision and recall rates, ensuring reliable lane detection even in complex traffic scenarios. Additionally, our model's real-time processing capabilities make it an ideal solution for implementation in AVs, enabling safer and more efficient navigation on roads.

show abstract

Computer Vision based Animal Collision Avoidance Framework for Autonomous Vehicles

Cited by 2 publications

References 9 publications

An Accurate and Fast Animal Species Detection System for Embedded Devices

An Accurate and Fast Animal Species Detection System for Embedded Devices

Mask R-CNN Approach to Real-Time Lane Detection for Autonomous Vehicles

Contact Info

Product

Resources

About