Abstract
A new hybrid vehicle detection scheme which integrates the Viola-Jones (V-J) and linear SVM classifier with HOG feature (HOG + SVM) methods is proposed for vehicle detection from low-altitude unmanned aerial vehicle (UAV) images. As both V-J and HOG + SVM are sensitive to on-road vehicles’ in-plane rotation, the proposed scheme first adopts a roadway orientation adjustment method, which rotates each UAV image to align the roads with the horizontal direction so the original V-J or HOG + SVM method can be directly applied to achieve fast detection and high accuracy. To address the issue of descending detection speed for V-J and HOG + SVM, the proposed scheme further develops an adaptive switching strategy which sophistically integrates V-J and HOG + SVM methods based on their different descending trends of detection speed to improve detection efficiency. A comprehensive evaluation shows that the switching strategy, combined with the road orientation adjustment method, can significantly improve the efficiency and effectiveness of the vehicle detection from UAV images. The results also show that the proposed vehicle detection method is competitive compared with other existing vehicle detection methods. Furthermore, since the proposed vehicle detection method can be performed on videos captured from moving UAV platforms without the need of image registration or additional road database, it has great potentials of field applications. Future research will be focusing on expanding the current method for detecting other transportation modes such as buses, trucks, motors, bicycles, and pedestrians.
Highlights
Unmanned aerial vehicles (UAVs) have been widely used in many fields, such as chemical vapour detection [1], nature conservation monitoring [2] and wildlife emergency response [3].unmanned aerial vehicle (UAV) hold great promise for transportation applications, as demonstrated by many transportation studies [4,5,6,7]
By incorporating the road orientation adjustment method, the proposed vehicle detection method will be insensitive to road orientation changes and can achieve high
ViBe, a universal background subtraction algorithm [31]; Frame difference [9]; Original V-J method [11]; Rotate each image every 20◦ from 0◦ to 180◦ and detect nine times using the original V-J method [14]; (5) Original V-J method combines with the proposed road orientation adjustment method only; (6) Original histogram of oriented gradients (HOG) + support vector machine (SVM) [12]; (7) Rotate each image every 20◦ from 0◦ to 180◦ and detect nine times using the original HOG + SVM
Summary
Unmanned aerial vehicles (UAVs) have been widely used in many fields, such as chemical vapour detection [1], nature conservation monitoring [2] and wildlife emergency response [3].UAVs hold great promise for transportation applications, as demonstrated by many transportation studies [4,5,6,7]. One important application of UAV technology for transportation is to enhance the traffic and emergency monitoring which has been serving as a backbone of Intelligent. Many of them applied some traditional methods, such as background subtraction, frame difference, optical flow, etc. Azevedo [8] applied a median-based background subtraction method to fast detect vehicles; Shastry and Schowengerdt [9] applied a frame difference method, combining with the image registration process to detect moving vehicles; and Yalcin [10] proposed a motion-based optical flow method to detect moving vehicles. Methods like frame difference, background subtraction and optical flow are sensitive to scene complexity have difficulties in detecting slow-moving or stopped vehicles when traffic is congested. Object detection algorithms are less sensitive to image noise, background motions and scene complexity, are more robust for vehicle detections from UAV videos
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
