Vision-based Edge-line Assisted Moving Vehicle Detection
13 Oct 2020-
TL;DR: Experimental results show that this low-cost vision-based moving vehicle detection method in streaming video can quickly and accurately identify moving vehicles on the road under a fixed viewing angle.
Abstract: The detection of vehicles has a very wide range of applications in public transportation monitoring and management. Especially the moving vehicles detection plays an indispensable role in modern Intelligent Transport System (ITS). A reliable mobile vehicle detection method can provide necessary guarantees for traffic safety, assist drivers or pedestrians to predict road conditions, such as traffic statistics in designated areas, driver blind spot assistance, thereby ensuring the safety of pedestrians, passengers and drivers. Recently, most of the deep learning research stays on object detection and classification in the image, few studies on moving object detection in streaming video. In this paper, a low-cost vision-based moving vehicle detection method in streaming video is introduced. In order to distinguish between moving vehicles and stationary vehicles, a robust lane line detection method is used to detect the lanes, thereby avoiding the interference of stationary targets and background factors on the outside of the lanes. Then select the pixels in the road area and fill the cut area with pure white pixels, pass this reorganized data matrix to a pre-trained vehicles detection deep neural network to get the moving vehicles information. Experimental results show that this method can quickly and accurately identify moving vehicles on the road under a fixed viewing angle.
Citations
More filters
01 Jan 2023
TL;DR: In this article , the authors proposed a method to make intelligent vehicles complete lateral lane change control, longitudinal overtaking control, and path planning of cloud computing network-connected road sections with the support of roadside-end fusion detection equipment.
Abstract: In this paper, we propose a method to make intelligent vehicles complete lateral lane change control, longitudinal overtaking control, and path planning of cloud computing network-connected road sections with the support of roadside-end fusion detection equipment. Also, we simulated the actual scenario of intelligent vehicles and vehicle detection equipment between the data communication and complete the communication test in the simulation scene of the sand table model. As for roadside detection equipment, different detection equipment is customized according to the traffic flow and traffic scenarios of real urban roads, sensors such as lidar and cameras are fused, and roadside information monitoring and vehicle scheduling are completed in combination with roadside units and computing platforms. As for the intelligent vehicle navigation module, lidar and binocular cameras are used for SLAM mapping, and the vehicle is accurately positioned with the help of the camera and Tag code. The simulation-to-real environment will improve the vehicle positioning accuracy through the WIFI module. For the intelligent vehicle control module, it provides real-time roadside information to the vehicle, which completes the tests of vehicle road driving, marker recognition, and traffic light detection in Gazebo by cameras and lidars.
References
More filters
Posted Content•
TL;DR: Faster R-CNN as discussed by the authors proposes a Region Proposal Network (RPN) to generate high-quality region proposals, which are used by Fast R-NN for detection.
Abstract: State-of-the-art object detection networks depend on region proposal algorithms to hypothesize object locations. Advances like SPPnet and Fast R-CNN have reduced the running time of these detection networks, exposing region proposal computation as a bottleneck. In this work, we introduce a Region Proposal Network (RPN) that shares full-image convolutional features with the detection network, thus enabling nearly cost-free region proposals. An RPN is a fully convolutional network that simultaneously predicts object bounds and objectness scores at each position. The RPN is trained end-to-end to generate high-quality region proposals, which are used by Fast R-CNN for detection. We further merge RPN and Fast R-CNN into a single network by sharing their convolutional features---using the recently popular terminology of neural networks with 'attention' mechanisms, the RPN component tells the unified network where to look. For the very deep VGG-16 model, our detection system has a frame rate of 5fps (including all steps) on a GPU, while achieving state-of-the-art object detection accuracy on PASCAL VOC 2007, 2012, and MS COCO datasets with only 300 proposals per image. In ILSVRC and COCO 2015 competitions, Faster R-CNN and RPN are the foundations of the 1st-place winning entries in several tracks. Code has been made publicly available.
23,183 citations
23 Jun 2014
TL;DR: RCNN as discussed by the authors combines CNNs with bottom-up region proposals to localize and segment objects, and when labeled training data is scarce, supervised pre-training for an auxiliary task, followed by domain-specific fine-tuning, yields a significant performance boost.
Abstract: Object detection performance, as measured on the canonical PASCAL VOC dataset, has plateaued in the last few years. The best-performing methods are complex ensemble systems that typically combine multiple low-level image features with high-level context. In this paper, we propose a simple and scalable detection algorithm that improves mean average precision (mAP) by more than 30% relative to the previous best result on VOC 2012 -- achieving a mAP of 53.3%. Our approach combines two key insights: (1) one can apply high-capacity convolutional neural networks (CNNs) to bottom-up region proposals in order to localize and segment objects and (2) when labeled training data is scarce, supervised pre-training for an auxiliary task, followed by domain-specific fine-tuning, yields a significant performance boost. Since we combine region proposals with CNNs, we call our method R-CNN: Regions with CNN features. We also present experiments that provide insight into what the network learns, revealing a rich hierarchy of image features. Source code for the complete system is available at http://www.cs.berkeley.edu/~rbg/rcnn.
21,729 citations
Posted Content•
[...]
TL;DR: This paper proposes a Fast Region-based Convolutional Network method (Fast R-CNN) for object detection that builds on previous work to efficiently classify object proposals using deep convolutional networks.
Abstract: This paper proposes a Fast Region-based Convolutional Network method (Fast R-CNN) for object detection. Fast R-CNN builds on previous work to efficiently classify object proposals using deep convolutional networks. Compared to previous work, Fast R-CNN employs several innovations to improve training and testing speed while also increasing detection accuracy. Fast R-CNN trains the very deep VGG16 network 9x faster than R-CNN, is 213x faster at test-time, and achieves a higher mAP on PASCAL VOC 2012. Compared to SPPnet, Fast R-CNN trains VGG16 3x faster, tests 10x faster, and is more accurate. Fast R-CNN is implemented in Python and C++ (using Caffe) and is available under the open-source MIT License at this https URL.
14,747 citations
Posted Content•
TL;DR: This paper proposes a simple and scalable detection algorithm that improves mean average precision (mAP) by more than 30% relative to the previous best result on VOC 2012 -- achieving a mAP of 53.3%.
Abstract: Object detection performance, as measured on the canonical PASCAL VOC dataset, has plateaued in the last few years. The best-performing methods are complex ensemble systems that typically combine multiple low-level image features with high-level context. In this paper, we propose a simple and scalable detection algorithm that improves mean average precision (mAP) by more than 30% relative to the previous best result on VOC 2012---achieving a mAP of 53.3%. Our approach combines two key insights: (1) one can apply high-capacity convolutional neural networks (CNNs) to bottom-up region proposals in order to localize and segment objects and (2) when labeled training data is scarce, supervised pre-training for an auxiliary task, followed by domain-specific fine-tuning, yields a significant performance boost. Since we combine region proposals with CNNs, we call our method R-CNN: Regions with CNN features. We also compare R-CNN to OverFeat, a recently proposed sliding-window detector based on a similar CNN architecture. We find that R-CNN outperforms OverFeat by a large margin on the 200-class ILSVRC2013 detection dataset. Source code for the complete system is available at this http URL.
13,081 citations
"Vision-based Edge-line Assisted Mov..." refers methods in this paper
...In 2010, the R-CNN [5] was proposed, the first step of this method is to run the image segmentation algorithm to obtain around 2000 cutting areas of different sizes, then place a bounding box on the different cutting areas and run the classifier, thereby avoiding the unnecessary calculation area and improves the efficiency of obstacle detection....
[...]
...Follow the idea of region-proposal method, R-CNN was developed to fast R-CNN [6] and faster R-CNN [7]....
[...]
Posted Content•
TL;DR: The authors present some updates to YOLO!
Abstract: We present some updates to YOLO! We made a bunch of little design changes to make it better. We also trained this new network that's pretty swell. It's a little bigger than last time but more accurate. It's still fast though, don't worry. At 320x320 YOLOv3 runs in 22 ms at 28.2 mAP, as accurate as SSD but three times faster. When we look at the old .5 IOU mAP detection metric YOLOv3 is quite good. It achieves 57.9 mAP@50 in 51 ms on a Titan X, compared to 57.5 mAP@50 in 198 ms by RetinaNet, similar performance but 3.8x faster. As always, all the code is online at this https URL
12,770 citations
"Vision-based Edge-line Assisted Mov..." refers methods in this paper
...YOLOv3 maintains the K-means clustering method adopted by YOLOv2 to obtain the size of the anchor, and increases its number to 9, in order to assign 3 different anchors to each scale of the feature map, so that the ability of the small objects detection is greatly improved....
[...]
...9d illustrates the YOLOV3 target detection process, Figure....
[...]
...Then, in the final step, YOLO target detection module is called to detect the moving target on the road, and the detected image is reassembled to get the completed result....
[...]
...The overall network structure of YOLOv3 is shown in Fig....
[...]
...YOLOv3 uses a network structure called Darknet-53, which deepens the network into 53 convolutional layers....
[...]