Vehicle Detection Technology Based on Cascading Classifiers of Multi-Feature Integration

2017 ◽  
Vol 31 (10) ◽  
pp. 1750032 ◽  
Author(s):  
Xinyu Hu ◽  
Xuhui Ye ◽  
Daode Zhang ◽  
Liangyi Wu
Keyword(s):  
Intelligent Transportation System ◽  
Gaussian Model ◽  
Vehicle Detection ◽  
Gaussian Mixture ◽  
Classification Performance ◽  
Feature Integration ◽  
Detection Technology ◽  
Single Feature ◽  
Scanning Area ◽  
Region Extraction

Vehicle detection, as an important technology for urban intelligent transportation system, is having attracted increasingly interests of researchers in recent years. For the time cost problem of traditional road vehicles testing approach, a moving region extraction method based on Gaussian model is used to reduce the scanning area of the window, exclude some background noise and improve test speed. For the problem of traditional single feature, relatively lower detection rate and lack of ability to adapt to complex environment, a method based on the combination of Haar-like and 2bitBP (2bit Binary Pattern) features is adopted. Feature integration method enhances the expression of features. As a result, the improved classification performance of classifiers enables it to be adapted to different traffic environment. Firstly, a Gaussian mixture model is established to detect moving targets in overall region and then the Haar-like and 2bitBP features extraction are carried out in the region. At the end the action of cascading classification on samples achieve the detection of moving vehicles. The experimental results show that the method is effective for vehicle detection.

scholarly journals Pretraining Convolutional Neural Networks for Image-Based Vehicle Classification

2018 ◽  
Vol 2018 ◽  
pp. 1-10 ◽  
Author(s):  
Yunfei Han ◽  
Tonghai Jiang ◽  
Yupeng Ma ◽  
Chunxiang Xu
Keyword(s):  
Neural Networks ◽  
Object Detection ◽  
Convolutional Neural Networks ◽  
Intelligent Transportation System ◽  
Vehicle Detection ◽  
Intelligent Transportation ◽  
Classification Performance ◽  
Vehicle Classification ◽  
Urban Computing ◽  
Detection Model

Vehicle detection and classification are very important for analysis of vehicle behavior in intelligent transportation system, urban computing, etc. In this paper, an approach based on convolutional neural networks (CNNs) has been applied for vehicle classification. In order to achieve a more accurate classification, we removed the unrelated background as much as possible based on a trained object detection model. In addition, an unsupervised pretraining approach has been introduced to better initialize CNNs parameters to enhance the classification performance. Through the data enhancement on manual labeled images, we got 2000 labeled images in each category of motorcycle, transporter, passenger, and others, with 1400 samples for training and 600 samples for testing. Then, we got 17395 unlabeled images for layer-wise unsupervised pretraining convolutional layers. A remarkable accuracy of 93.50% is obtained, demonstrating the high classification potential of our approach.

Evaluation of Commercial Truck Parking Detection for Rest Areas

2018 ◽  
Vol 2672 (9) ◽  
pp. 141-151
Author(s):  
Wei Sun ◽  
Ethan Stoop ◽  
Scott S. Washburn
Keyword(s):  
Software Tool ◽  
Vehicle Detection ◽  
Ground Truth ◽  
Early Morning ◽  
Video Data ◽  
Sensor Data ◽  
Parking Space ◽  
Ground Truth Data ◽  
Detection Technology ◽  
Rest Areas

Florida’s interstate rest areas are heavily utilized by commercial trucks for overnight parking. Many of these rest areas regularly experience 100% utilization of available commercial truck parking spaces during the evening and early-morning hours. Being able to communicate availability of commercial truck parking space to drivers in advance of arriving at a rest area would reduce unnecessary stops at full rest areas as well as driver anxiety. In order to do this, it is critical to implement a vehicle detection technology to reflect the parking status of the rest area correctly. The objective of this project was to evaluate three different wireless in-pavement vehicle detection technologies as applied to commercial truck parking at interstate rest areas. This paper mainly focuses on the following aspects: (a) accuracy of the vehicle detection in parking spaces, (b) installation, setup, and maintenance of the vehicle detection technology, and (c) truck parking trends at the rest area study site. The final project report includes a more detailed summary of the evaluation. The research team recorded video of the rest areas as the ground-truth data and developed a software tool to compare the video data with the parking sensor data. Two accuracy tests (event accuracy and occupancy accuracy) were conducted to evaluate each sensor’s ability to reflect the status of each parking space correctly. Overall, it was found that all three technologies performed well, with accuracy rates of 95% or better for both tests. This result suggests that, for implementation, pricing, and/or maintenance issues may be more significant factors for the choice of technology.

Evaluation of the Accuracy of Stop Bar Video Vehicle Detection at Signalized Intersections

2005 ◽  
Vol 1925 (1) ◽  
pp. 134-145 ◽  
Author(s):  
Avery Rhodes ◽  
Darcy M. Bullock ◽  
James Sturdevant ◽  
Zachary Clark ◽  
David G. Candey
Keyword(s):  
Video Recording ◽  
Evaluation Criteria ◽  
Vehicle Detection ◽  
Signalized Intersections ◽  
Human Observer ◽  
Loop Detectors ◽  
Detection Technology ◽  
Video Detection ◽  
Camera Location ◽  
Marginal Improvement

Many U.S. agencies have adopted video vehicle detection technology as an alternative to inductive loops. Although many product evaluations have been performed, the majority of these evaluations concentrated on freeway applications in which speed and volume were the primary evaluation criteria. At an actuated intersection, the metrics of speed and volume do not necessarily represent how well a device will operate as a presence detector. Video detection at signalized intersections was evaluated at a test intersection in Indiana. Cameras on all approaches were located at the optimal camera position recommended by the vendors, approximately 60 ft from the strain pole. Two additional cameras were located on each mast arm at slightly less optimal positions, 36 and 48 ft from the strain pole. Traditional inductive loops were also available at the intersection and were used to provide baseline data to screen for discrepancies. Each time the detectors were not in agreement, a discrepancy was noted. A digital video recording was later viewed by a human observer to determine whether the video detector or the loop detector was in error. An analysis of the data showed that video detection was found to produce statistically significantly more false detections and missed detections than the loop detectors on most phases. A small incremental increase in performance was observed when the camera was mounted at 60 ft rather than 36 ft on two of the approaches, but this marginal improvement likely does not justify the additional expense of mast arm, pole, and pole foundation associated with this camera location.

scholarly journals Recent Progress in Voice-Based Sasang Constitutional Medicine: Improving Stability of Diagnosis

2013 ◽  
Vol 2013 ◽  
pp. 1-7 ◽  
Author(s):  
Jun-Su Jang ◽  
Young-Su Kim ◽  
Boncho Ku ◽  
Jong Yeol Kim
Keyword(s):  
Gaussian Mixture ◽  
Voice Signal ◽  
Single Feature ◽  
Sasang Constitution ◽  
Diagnosis Accuracy ◽  
Good Repeatability ◽  
Diagnosis Model ◽  
Sasang Constitutional Medicine ◽  
Feature Values ◽  
Constitutional Medicine

Sasang constitutional medicine is a unique form of tailored medicine in traditional Korean medicine. Voice features have been regarded as an important cue to diagnose Sasang constitution types. Many studies tried to extract quantitative voice features and standardize diagnosis methods; however, they had flaws, such as unstable voice features which vary a lot for the same individual, limited data collected from only few sites, and low diagnosis accuracy. In this paper, we propose a stable diagnosis model that has a good repeatability for the same individual. None of the past studies evaluated the repeatability of their diagnosis models. Although many previous studies used voice features calculated by averaging feature values from all valid frames in monotonic utterance like vowels, we analyse every single feature value from each frame of a sentence voice signal. Gaussian mixture model is employed to deal with a lot of voice features from each frame. Total 15 Gaussian models are used to represent voice characteristics for each constitution. To evaluate repeatability of the proposed diagnosis model, we introduce a test dataset consisting of 10 individuals’ voice recordings with 50 recordings per each individual. Our result shows that the proposed method has better repeatability than the previous study which used averaged features from vowels and the sentence.

Detection of Moving Objects Based on Mixture Gaussian Model

2014 ◽  
Vol 1039 ◽  
pp. 274-279
Author(s):  
Guang Hua Chen ◽  
Gui Zhi Sheng
Keyword(s):  
Moving Objects ◽  
Video Segmentation ◽  
Color Space ◽  
Gaussian Model ◽  
Gaussian Mixture ◽  
Connected Components ◽  
Shadow Removal ◽  
Em Algorithms ◽  
Learning Rates ◽  
Components Analysis

The paper proposes an improved adaptive Gaussian mixture model (GMM) approach with online EM algorithms for updating, which solves the video segmentation problems carried by busy environment and illumination change. Different learning rates are set for foreground district and background district respectively, which improves the convergence speed of background model. A shadow removal scheme is also introduced for extracting complete moving objects. It is based on brightness distortion and chromaticity distortion in RGB color space. Morphological filtering and connected components analysis algorithm are also introduced to process the result of background subtraction. The experiment results show that the improved GMM has good accuracy and high adaptability in video segmentation. It can extract a complete and clear moving object when it is incorporated with shadow removal.

Sign in / Sign up

Export Citation Format

Share Document