Deep-Sea Organisms Tracking Using Dehazing and Deep Learning

Deep-sea organism automatic tracking has rarely been studied because of a lack of training data. However, it is extremely important for underwater robots to recognize and to predict the behavior of organisms. In this paper, we first develop a method for underwater real-time recognition and tracking of multi-objects, which we call “You Only Look Once: YOLO”. This method provides us with a very fast and accurate tracker. At first, we remove the haze, which is caused by the turbidity of the water from a captured image. After that, we apply YOLO to allow recognition and tracking of marine organisms, which include shrimp, squid, crab and shark. The experiments demonstrate that our developed system shows satisfactory performance.

[1]  Jenq-Neng Hwang,et al.  Underwater Fish Tracking for Moving Cameras Based on Deformable Multiple Kernels , 2016, IEEE Transactions on Systems, Man, and Cybernetics: Systems.

[2]  Huimin Lu,et al.  FDCNet: filtering deep convolutional network for marine organism classification , 2018, Multimedia Tools and Applications.

[3]  Takeo Kanade,et al.  An Iterative Image Registration Technique with an Application to Stereo Vision , 1981, IJCAI.

[4]  Yang Bai,et al.  Underwater object tracking using electrical impedance tomography , 2012, 2012 IEEE/RSJ International Conference on Intelligent Robots and Systems.

[5]  Christof Koch,et al.  Detection and tracking of objects in underwater video , 2004, CVPR 2004.

[6]  Huchuan Lu,et al.  Kernel collaborative face recognition , 2015, Pattern Recognit..

[7]  Huimin Lu,et al.  Contrast enhancement for images in turbid water. , 2015, Journal of the Optical Society of America. A, Optics, image science, and vision.

[8]  Jiri Matas,et al.  Forward-Backward Error: Automatic Detection of Tracking Failures , 2010, 2010 20th International Conference on Pattern Recognition.

[9]  Judea Pearl,et al.  Some Recent Results in Heuristic Search Theory , 1984, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[10]  Xuelong Li,et al.  Two dimensional principal components of natural images and its application , 2011, Neurocomputing.

[11]  Huimin Lu,et al.  Underwater image dehazing using joint trilateral filter , 2014, Comput. Electr. Eng..

[12]  Jun Luo,et al.  Detection and Tracking of Underwater Object Based on Forward-Scan Sonar , 2012, ICIRA.

[13]  Ali Farhadi,et al.  You Only Look Once: Unified, Real-Time Object Detection , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[14]  Nikola Miskovic,et al.  Underwater Object Tracking Using Sonar and USBL Measurements , 2016, J. Sensors.

[15]  Huimin Lu,et al.  Brain Intelligence: Go beyond Artificial Intelligence , 2017, Mobile Networks and Applications.

[16]  Hyun Myung,et al.  Vision-based object detection and tracking for autonomous navigation of underwater robots , 2012 .

[17]  Huimin Lu,et al.  Underwater image de-scattering and classification by deep neural network , 2016, Comput. Electr. Eng..

[18]  Huimin Lu,et al.  Motor Anomaly Detection for Unmanned Aerial Vehicles Using Reinforcement Learning , 2018, IEEE Internet of Things Journal.

[19]  Ali Farhadi,et al.  YOLO9000: Better, Faster, Stronger , 2016, 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[20]  Zhenbang Gong,et al.  Underwater object detection and tracking based on multi-beam sonar image processing , 2013, 2013 IEEE International Conference on Robotics and Biomimetics (ROBIO).

[21]  Huimin Lu,et al.  Turbidity Underwater Image Restoration Using Spectral Properties and Light Compensation , 2016, IEICE Trans. Inf. Syst..

[22]  Huimin Lu,et al.  Low illumination underwater light field images reconstruction using deep convolutional neural networks , 2018, Future Gener. Comput. Syst..