Zooplankton are quite significant to the ocean ecosystem for stabilizing balance of the ecosystem and keeping the earth running normally. Considering the significance of zooplantkon, research about zooplankton has caught more and more attentions. And zooplankton recognition has shown great potential for science studies and mearsuring applications. However, manual recognition on zooplankton is labour-intensive and time-consuming, and requires professional knowledge and experiences, which can not scale to large-scale studies. Deep learning approach has achieved remarkable performance in a number of object recognition benchmarks, often achieveing the current best performance on detection or classification tasks and the method demonstrates very promising and plausible results in many applications. In this paper, we explore a deep learning architecture: ZooplanktoNet to classify zoolankton automatically and effectively. The deep network is characterized by capturing more general and representative features than previous predefined feature extraction algorithms in challenging classification. Also, we incorporate some data augmentation to aim at reducing the overfitting for lacking of zooplankton images. And we decide the zooplankton class according to the highest score in the final predictions of ZooplanktoNet. Experimental results demonstrate that ZooplanktoNet can solve the problem effectively with accuracy of 93.7% in zooplankton classification.978-1-4673-9724-7/16/$31.00 ©2016 IEEE

Moving object detection is the basis of object tracking and classification in computer vision, which has been applied to underwater robots to execute underwater missions and do marine ecological research. The complicated scene and poor lighting condition in underwater environment usually make moving object detection difficult. To solve above problems and to detect moving object from underwater video, we propose an approach combining background subtraction and three-frame difference. In this method, firstly we detect moving object pixels by background subtraction and three-frame difference perspectively. Next, we perform "AND" operation on the results of background subtraction and three-frame difference, background subtraction provides the object information to supplement the incomplete information detected from three-frame difference. Lastly, morphology processing is utilized on the result to remove the noise caused by non-static objects in background. This method shows an effective and reliable performance in detecting moving object from underwater video.

No abstract

No abstract

Zooplankton are the key components of marine food webs. The abundance of it influences the ocean ecological balance. To efficiently monitor species richness of zooplankton and protect marine environment, marine biologists and computer vision experts started to research automated zooplankton classification system with computer vision technologies. Most current research focuses on achieving high classification accuracy. In this paper, we propose a new system based on multi features combination to enhance the zooplankton classification performance. In our system, the geometric and grayscale features, Local Binary Patterns features, and Inner-distance Shape Context features are extracted as low-level features. According to the properties of machine learning algorithms, an appropriate algorithm is chosen to generate middle-level features by processing all kinds of lowlevel features. After that, we concatenate middle-level features and apply Support Vector Machine to get the final classifier. By combining different types of features, the system we proposed can capture richer biomorphic information than those with a few features. And the experimental results also show that our system achieves better classification performance.