scholarly journals Traffic Lights Detection and Recognition Method Based on the Improved YOLOv4 Algorithm

Sensors ◽  
2021 ◽  
Vol 22 (1) ◽  
pp. 200
Author(s):  
Qingyan Wang ◽  
Qi Zhang ◽  
Xintao Liang ◽  
Yujing Wang ◽  
Changyue Zhou ◽  
...  

For facing of the problems caused by the YOLOv4 algorithm’s insensitivity to small objects and low detection precision in traffic light detection and recognition, the Improved YOLOv4 algorithm is investigated in the paper using the shallow feature enhancement mechanism and the bounding box uncertainty prediction mechanism. The shallow feature enhancement mechanism is used to extract features from the network and improve the network’s ability to locate small objects and color resolution by merging two shallow features at different stages with the high-level semantic features obtained after two rounds of upsampling. Uncertainty is introduced in the bounding box prediction mechanism to improve the reliability of the prediction of the bounding box by modeling the output coordinates of the prediction bounding box and adding the Gaussian model to calculate the uncertainty of the coordinate information. The LISA traffic light data set is used to perform detection and recognition experiments separately. The Improved YOLOv4 algorithm is shown to have a high effectiveness in enhancing the detection and recognition precision of traffic lights. In the detection experiment, the area under the PR curve value of the Improved YOLOv4 algorithm is found to be 97.58%, which represents an increase of 7.09% in comparison to the 90.49% score gained in the Vision for Intelligent Vehicles and Applications Challenge Competition. In the recognition experiment, the mean average precision of the Improved YOLOv4 algorithm is 82.15%, which is 2.86% higher than that of the original YOLOv4 algorithm. The Improved YOLOv4 algorithm shows remarkable advantages as a robust and practical method for use in the real-time detection and recognition of traffic signal lights.

2021 ◽  
Vol 11 (17) ◽  
pp. 8066
Author(s):  
Tien-Wen Yeh ◽  
Huei-Yung Lin ◽  
Chin-Chen Chang

We present a traffic light detection and recognition approach for traffic lights that utilizes convolutional neural networks. We also introduce a technique for identifying arrow signal lights in multiple urban traffic environments. For detection, we use map data and two different focal length cameras for traffic light detection at various distances. For recognition, we propose a new algorithm that combines object detection and classification to recognize the light state classes of traffic lights. Furthermore, we use a unified network by sharing features to decrease computation time. The results reveal that the proposed approach enables high-performance traffic light detection and recognition.


2020 ◽  
Vol 9 (1) ◽  
pp. 2526-2534

This paper principally combines ideas of laptop vision, machine learning and deep learning for correct detection of traffic lights and their classifications. It checks for each circular and arrow stoplight cases. Color filtering and blob discover ion area unit principally to detect the candidates (traffic lights) [6]. Then, a PCA network is employed as a multiclass classifier which provides the result sporadically. MOT will used for more trailing method and prediction filters out false positives. Sometimes, vote theme can even be used rather than MOT. This method will be simply fitted into ADAS vehicles once hardware thinks about. Recognition is as vital as detective work the traffic lights. While not recognition, no full data will be transmitted [2]. Many complicated TLR’s will give advance functions like observing the most the most for a specific route (when there's quite one) and the way removed from the driving force [3]. Deep learning is additionally one among the rising techniques for analysis areas [7]. Object detection comes as associate integral a part of laptop vision. Object detection will be best utilized in create estimation, vehicle detection, police work etc. In detection algorithms, we tend to incline to draw a bounding box round the object of interest to find it among the image. Also, the drawing of the bounding box isn't distinctive and might hyperbolically looking on the need [9].


2021 ◽  
Vol 8 (2) ◽  
pp. 15-19
Author(s):  
Julkar Nine ◽  
Rahul Mathavan

Traffic light detection and back-light recognition are essential research topics in the area of intelligent vehicles because they avoid vehicle collision and provide driver safety. Improved detection and semantic clarity may aid in the prevention of traffic accidents by self-driving cars at crowded junctions, thus improving overall driving safety. Complex traffic situations, on the other hand, make it more difficult for algorithms to identify and recognize objects. The latest state-of-the-art algorithms based on Deep Learning and Computer Vision are successfully addressing the majority of real-time problems for autonomous driving, such as detecting traffic signals, traffic signs, and pedestrians. We propose a combination of deep learning and image processing methods while using the MobileNetSSD (deep neural network architecture) model with transfer learning for real-time detection and identification of traffic lights and back-light. This inference model is obtained from frameworks such as Tensor-Flow and Tensor-Flow Lite which is trained on the COCO data. This study investigates the feasibility of executing object detection on the Raspberry Pi 3B+, a widely used embedded computing board. The algorithm’s performance is measured in terms of frames per second (FPS), accuracy, and inference time.


2012 ◽  
Vol 151 ◽  
pp. 510-513 ◽  
Author(s):  
Yu Peng Yao ◽  
Ying Shi ◽  
Ji You Fei

Configuration technology is a new technology for monitoring in the current society; it is the result of the development of computer control technology. To traffic light control system, it is to combine the use of configuration technology and procedures related to PLC, and through software simulation and traffic lights light changes, traffic light control system could achieve the monitoring problem, and if the system is in good condition, its application can save a lot of labor powers and materials.


2021 ◽  
Vol 116 (1) ◽  
pp. 299-304
Author(s):  
Assel Aliyadynovna Sailau

The number of vehicles on the roads of Almaty, Kazakhstan is growing from year to year. This brings about an increasing intensity and density of traffic flows in the streets which leads to congestion, decreasing speed of the traffic flow, increasing environmental pollution caused by car emissions, and which can potentially lead to the road traffic accidents (RTA), including fatalities. While the number of injuries grows up mainly due to drivers’ non-compliance with the speed limit, the environmental pollution is caused by longer traffic jams. Therefore, to reduce the level of road traffic injuries and emissions into the environment it is necessary to ensure the uniform movement of traffic flows in cities. Currently, one of the effective ways to do it is the use of transport telematics systems, in particular, control systems for road signs, road boards and traffic lights. The paper presents an analysis of existing systems and methods of traffic light regulation. The  analyses of the systems and methods are based on the use of homogeneous data, that is the data on standard parameters of traffic flows. The need in collecting and analyzing additional semi-structured data on the factors that have a significant impact on the traffic flows parameters in cities is shown as well. The work is dedicated to solving the problem of analysis and forecast of traffic flows in the city of Almaty, Kazakhstan. GPS data on the location of individual vehicles is used as the initial data for solving this problem. By projecting the obtained information onto the graph of the city's transport network, as well as using additional filtering, it is possible to obtain an estimate of individual parameters of traffic flows. These parameters are used for short-term forecast of the changes in the city's transport network.


Author(s):  
Muhammad Rusyadi Ramli ◽  
Riesa Krisna Astuti Sakir ◽  
Dong-Seong Kim

This paper presents fog-based intelligent transportation systems (ITS) architecture for traffic light optimization. Specifically, each intersection consists of traffic lights equipped with a fog node. The roadside unit (RSU) node is deployed to monitor the traffic condition and transmit it to the fog node. The traffic light center (TLC) is used to collect the traffic condition from the fog nodes of all intersections. In this work, two traffic light optimization problems are addressed where each problem will be processed either on fog node or TLC according to their requirements. First, the high latency for the vehicle to decide the dilemma zone is addressed. In the dilemma zone, the vehicle may hesitate whether to accelerate or decelerate that can lead to traffic accidents if the decision is not taken quickly. This first problem is processed on the fog node since it requires a real-time process to accomplish. Second, the proposed architecture aims each intersection aware of its adjacent traffic condition. Thus, the TLC is used to estimate the total incoming number of vehicles based on the gathered information from all fog nodes of each intersection. The results show that the proposed fog-based ITS architecture has better performance in terms of network latency compared to the existing solution in which relies only on TLC.


Information ◽  
2018 ◽  
Vol 9 (11) ◽  
pp. 278 ◽  
Author(s):  
Thiago Almeida ◽  
Hendrik Macedo ◽  
Leonardo Matos ◽  
Nathanael Vasconcelos

Traffic light detection and recognition (TLR) research has grown every year. In addition, Machine Learning (ML) has been largely used not only in traffic light research but in every field where it is useful and possible to generalize data and automatize human behavior. ML algorithms require a large amount of data to work properly and, thus, a lot of computational power is required to analyze the data. We argue that expert knowledge should be used to decrease the burden of collecting a huge amount of data for ML tasks. In this paper, we show how such kind of knowledge was used to reduce the amount of data and improve the accuracy rate for traffic light detection and recognition. Results show an improvement in the accuracy rate around 15%. The paper also proposes a TLR device prototype using both camera and processing unit of a smartphone which can be used as a driver assistance. To validate such layout prototype, a dataset was built and used to test an ML model based on adaptive background suppression filter (AdaBSF) and Support Vector Machines (SVMs). Results show 100% precision rate and recall of 65%.


2020 ◽  
Vol 8 (6) ◽  
pp. 3228-3231

Intelligent Transport System (ITS) is blooming worldwide. The Traditional Traffic management system is a tedious process and it requires huge man power, to overcome this we have proposed an automatic Traffic monitoring system that has effective fleet management. The current transportation system at intersections and junctions has Traffic Lights with Fixed durations which increase the unnecessary staying time which intern harms the environment. An Adaptive traffic light control is implemented using SUMO simulator, that changes the duration of Green and Red light according to the traffic flow. This is an effective and efficient way to reduce the Traffic congestion. The traffic congestion is determined by taking the object count using deep learning approach (Convolutional Neural Network).


2020 ◽  
Vol 26 (2) ◽  
pp. 192-201
Author(s):  
Sri Redjeki Pudjaprasetya ◽  
Dear Michiko Noor

Traffic management of intersections is an important factor that can determine traffic density at the intersection, as well as its surrounding. Long traffic queues we encounter in daily life, were often caused by ineffectiveness of traffic lights management of the cross sections.In this article, an analytic study of traffic light management of a four-leg intersection, based on the kinematic LWR model, was presented. Comparison was based on observing the end of queues over three cycles of red-green lights, under the assumption of a constant traffic flux. On every leg of the intersection, the end of the queues were obtained from characteristic lines of the shock waves.From these observations, the three phase regulation was preferred over the four-phase one. Finally, a case study of Taman Sari - Baltos intersection located in Bandung City, Indonesia, was discussed. Parameter values used in these simulations were obtained from direct observation. 


Sign in / Sign up

Export Citation Format

Share Document