Vehicle Detection from Remote Sensing Image Based on Superpixel Segmentation and Image Enhancement

Automatic vehicle detection from high-resolution remote sensing image is a challenging topic. While there have been some studies on this topic in recent years, a fast and robust approach is yet to be found, especially when facing the scenario of low color contrast. In this paper, a new vehicle detection approach is proposed. First, superpixel-based segmentation is used to identify potential vehicle regions to speed up the detection and improve the accuracy. Then, an image enhancement method is also proposed, which greatly improves the segmentation results. Support vector machine is used to classification with features extracted by HOG descriptor. According to the experiments, by combining with superpixel segmentation and the image enhancement, the speed of the vehicle detection is improved with approximately an order of magnitude. Also, in case of low contrast remote sensing images, the detection accuracy can be also greatly improved, with much less false positives and false negatives.

[1]  Li-Der Chou,et al.  Intersection-Based Routing Protocol for VANETs , 2011, Wirel. Pers. Commun..

[2]  Hong Qiao,et al.  Airborne moving vehicle detection for video surveillance of urban traffic , 2009, 2009 IEEE Intelligent Vehicles Symposium.

[3]  Pascal Fua,et al.  SLIC Superpixels Compared to State-of-the-Art Superpixel Methods , 2012, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[4]  Jae-Young Choi,et al.  Vehicle Detection from Aerial Images Using Local Shape Information , 2009, PSIVT.

[5]  Stefan Hinz,et al.  Vehicle Detection in Aerial Images Using Generic Features, Grouping, and Context , 2001, DAGM-Symposium.

[6]  Luc Van Gool,et al.  SEEDS: Superpixels Extracted via Energy-Driven Sampling , 2012, ECCV.

[7]  Deren Li,et al.  Histogram of oriented gradient detector with color-invariant gradients in Gaussian color space , 2010 .

[8]  Supun Samarasekera,et al.  Aerial video surveillance and exploitation , 2001, Proc. IEEE.

[9]  Christopher J. C. Burges,et al.  A Tutorial on Support Vector Machines for Pattern Recognition , 1998, Data Mining and Knowledge Discovery.

[10]  Suman Srinivasan,et al.  Airborne traffic surveillance systems: video surveillance of highway traffic , 2004, VSSN '04.

[11]  Rama Chellappa,et al.  Entropy rate superpixel segmentation , 2011, CVPR 2011.

[12]  Hsu-Yung Cheng,et al.  Vehicle Detection in Aerial Surveillance Using Dynamic Bayesian Networks , 2012, IEEE Transactions on Image Processing.

[13]  Luc Van Gool,et al.  SEEDS: Superpixels Extracted Via Energy-Driven Sampling , 2012, International Journal of Computer Vision.