Vehicle detection using Gaussian mixture model for infrared orientation-code image

Hirata, Naru; Okuda, Hiroyuki; Seki, Masanori; Hashimoto, Manabu

doi:10.1117/12.648770

Cited by 2 publications

(1 citation statement)

References 6 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the 1960s, researchers began to conduct preliminary research on vehicle target detection [ 1 , 2 ]. It is a wide consensus that vehicle target detection is the basis of high-end technologies such as traffic big data platform construction, vehicle information management, and automatic driving [ 3 , 4 ].…”

Section: Introductionmentioning

confidence: 99%

Research on Deep Learning Automatic Vehicle Recognition Algorithm Based on RES-YOLO Model

Wang

Huang

et al. 2022

Sensors

View full text Add to dashboard Cite

With the introduction of concepts such as ubiquitous mapping, mapping-related technologies are gradually applied in autonomous driving and target recognition. There are many problems in vision measurement and remote sensing, such as difficulty in automatic vehicle discrimination, high missing rates under multiple vehicle targets, and sensitivity to the external environment. This paper proposes an improved RES-YOLO detection algorithm to solve these problems and applies it to the automatic detection of vehicle targets. Specifically, this paper improves the detection effect of the traditional YOLO algorithm by selecting optimized feature networks and constructing adaptive loss functions. The BDD100K data set was used for training and verification. Additionally, the optimized YOLO deep learning vehicle detection model is obtained and compared with recent advanced target recognition algorithms. Experimental results show that the proposed algorithm can automatically identify multiple vehicle targets effectively and can significantly reduce missing and false rates, with the local optimal accuracy of up to 95% and the average accuracy above 86% under large data volume detection. The average accuracy of our algorithm is higher than all five other algorithms including the latest SSD and Faster-RCNN. In average accuracy, the RES-YOLO algorithm for small data volume and large data volume is 1.0% and 1.7% higher than the original YOLO. In addition, the training time is shortened by 7.3% compared with the original algorithm. The network is then tested with five types of local measured vehicle data sets and shows satisfactory recognition accuracy under different interference backgrounds. In short, the method in this paper can complete the task of vehicle target detection under different environmental interferences.

show abstract

Section: Introductionmentioning

confidence: 99%

Research on Deep Learning Automatic Vehicle Recognition Algorithm Based on RES-YOLO Model

Wang

Huang

et al. 2022

Sensors

View full text Add to dashboard Cite

show abstract

A Flexible Algorithm for Detecting Challenging Moving Objects in Real-Time within IR Video Sequences

2017

View full text Add to dashboard Cite

Real-time detecting moving objects in infrared video sequences may be particularly challenging because of the characteristics of the objects, such as their size, contrast, velocity and trajectory. Many proposed algorithms achieve good performances but only in the presence of some specific kinds of objects, or by neglecting the computational time, becoming unsuitable for real-time applications. To obtain more flexibility in different situations, we developed an algorithm capable of successfully dealing with small and large objects, slow and fast objects, even if subjected to unusual movements, and poorly-contrasted objects. The algorithm is also capable to handle the contemporary presence of multiple objects within the scene and to work in real-time even using cheap hardware. The implemented strategy is based on a fast but accurate background estimation and rejection, performed pixel by pixel and updated frame by frame, which is robust to possible background intensity changes and to noise. A control routine prevents the estimation from being biased by the transit of moving objects, while two noise-adaptive thresholding stages, respectively, drive the estimation control and allow extracting moving objects after the background removal, leading to the desired detection map. For each step, attention has been paid to develop computationally light solution to achieve the real-time requirement. The algorithm has been tested on a database of infrared video sequences, obtaining promising results against different kinds of challenging moving objects and outperforming other commonly adopted solutions. Its effectiveness in terms of detection performance, flexibility and computational time make the algorithm particularly suitable for real-time applications such as intrusion monitoring, activity control and detection of approaching objects, which are fundamental task in the emerging research area of Smart City.

show abstract

Vehicle detection using Gaussian mixture model for infrared orientation-code image

Cited by 2 publications

References 6 publications

Research on Deep Learning Automatic Vehicle Recognition Algorithm Based on RES-YOLO Model

Research on Deep Learning Automatic Vehicle Recognition Algorithm Based on RES-YOLO Model

A Flexible Algorithm for Detecting Challenging Moving Objects in Real-Time within IR Video Sequences

Contact Info

Product

Resources

About