Shark detection in uncontrolled environment is a challenging problem that has not been dug into very deeply. This thesis presents a fast and effective detection system for submerged shark such as the white shark using image identification. The proposed system adopts an unmanned aerial vehicle (UAV) equipping with a low resolution camera for coast surveillance in low altitude. One of the goals of this research is to develop a technique to balances computational cost and detection effectiveness. The image identification is trained from real datasets and is applied to a 2.5-meter and 0.945 m2 shark model. The Haar feature-based cascade classifier is used to detect regions of interest (ROI). Then, it extracts some features, such as area, shape, time consistency, to classify whether or not the area contains a shark. The proposed system is tested in two different uncontrolled areas in Kenting coast, both presenting contrasted conditions. The technique used reached an average of 19 frames/second based on different altitudes of UAV experiments from 8 to 22 meters above sea level in static and dynamic detection. The system achieved a true detection’s average of 99.5% (correct classification) and the mean score on total false positive detection is 3.85%. The detection rate varies with the altitude and the weather conditions which proves the effectiveness of building an altitude-based system. The experiments show encouraging results that can be improved to reach a lower false positive detection rate. A more efficient ROI detector can further improve the computational cost and thus allow to process additional features such as side fins position in real time detection.

[1] D. Gibbins and D. A. Gray, D. Dempsey. “Classifying ships using low resolution maritime radar”, fifth International Symposium on Signal Processing and its Applications, ISSPA '99, 22-25 August, 1999, volume 2, pp. 325-329.
[2] X. Xie, J. Hong and S. Xiao, “Effective Method for Moving Objects Detection on Sea Surface”, 2008 International Conference on Computer Science and Software Engineering, Volume 6.
[3] C. Zhang, Y. Qiao, E. Fallon, C. Xu, “An Improved CamShift Algorithm for Target Tracking in Video Surveillance”, The 9th IT & T Conference, Dublin Institute of Technology, Dublin, Ireland, 22nd -23rd October, 2009.
[4] W. D. Robbins, V. M. Peddemors, S. J. Kennelly, M. C. Ives, “Experimental Evaluation of Shark Detection Rates by Aerial Observers”, PloS one, 3rd February 2014.
[5] G. Arceneaux IV, A. K. Schedin, A. J. W. White, “Real time face and facial feature detection and tracking”, Journal of Network and Computer Applications Volume 33, Issue 4, July 2010, pp. 447–466.
[6] B. Hughes, T. Burghardt, “Automated Identification of Individual Great White Sharks from Unrestricted Fin Imagery”, Proceedings of the 26th British Machine Vision Conference (BMVC), September 2015, pp. 92.1-92.14.
[7] P. Viola and M. Jones, “Rapid Object Detection using a Boosted Cascade of Simple Features”, 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001, volume 2, pp. I-511 – I-519.
[8] G. R. Bradski, “Computer Vision Face Tracking for Use in a Perceptual User Interface”, Fourth IEEE Workshop on Applications of Computer Vision, 1998. WACV '98, pp. 214,219, 19-21 October 1998.
[9] S. Gururatsakul, “Pattern recognition techniques for the identification of white sharks in aerial image sequences”, Doctoral thesis, Information Technology Adelaide, South Australia, Australia, 2011.
[10] S. Gururatsakul, D. Gibbins, D. Kearney, I. Lee, "Shark Detection using Optical Image Data from a Mobile Aerial Platform", 25th International Conference Image and Vision Computing New Zealand (IVCNZ 2010), pp, 8-9 November 2010.
[11] S. Gururatsakul, D. Gibbins, D. Kearney, "A Simple Deformable Model for Shark Recognition", Canadian Conference on Computer and Robot Vision (CRV 2011), vol., no., pp.234-241, 25-27 May 2011.
[12] G. T. Shrivakshan, “An Analysis of SOBEL and GABOR Image Filters for Identifying Fish”. Proceedings of the 2013 International Conference on Pattern Recognition, Informatics and Mobile Engineering (PRIME) February 21-22.
[13] L. Mejias, G. Duclos, A. Hodgson, F. D. Maire, “Automated marine mammal detection from aerial imagery”. IEEE/MTS, Proceedings of OCEANS 2013.
[14] F. S. Leira, T. A. Johansen & T. I. Fossen, “Automatic Detection, Classification and Tracking of Objects in the Ocean Surface from UAVs Using a Thermal Camera”, 7-14 March 2015 IEEE Aerospace Conference.
[15] J. S. Marques, A. Bernardino & G. Cruz, M. Bento, “An Algorithm for the Detection of Vessels in Aerial Images”, 2014 11th IEEE International Conference on Advanced Video and Signal Based Surveillance (AVSS) pp. 295-301.
[16] Y. Yu, S. Bi and Y. Mo, “Real-time gesture recognition system based on Camshift algorithm and Haar-like feature”, 19-22 June 2016 IEEE International Conference on Cyber Technology in Automation, Control, and Intelligent Systems (CYBER) pp. 337-343.
[17] Y. Freund, R. E. Schapire, "A short introduction to boosting", Journal of Japanese Society for Artificial Intelligence, vol. 14, no. 5, pp. 771-780, 1999.
[18] K. Fukunaga, L. D. Hostetler, "The Estimation of the Gradient of a Density Function, with Applications in Pattern Recognition", IEEE Transactions on Information Theory, 21 (1), January 1975, pp. 32–40.
[19] T. Ojala, M. Pietikïnen, D. Harwood, “A comparative study of texture measures a with classification based on feature distributions”, Pattern Recognition 29(1), January 1996, pp. 51–59.
[20] N. Dalal and B. Triggs, “Histograms of oriented gradients for human detection”, IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05), Volume 2, 20-25 June 2005, pp. 886-894.