Three-dimensional location of target fish by monocular infrared imaging sensor based on a L-z correlation model

文献类型： 外文期刊

第一作者： Lin, Kai

作者： Lin, Kai;Zhou, Chao;Xu, Daming;Guo, Qiang;Yang, Xinting;Sun, Chuanheng;Lin, Kai;Zhou, Chao;Xu, Daming;Guo, Qiang;Yang, Xinting;Sun, Chuanheng;Lin, Kai;Zhou, Chao;Xu, Daming;Guo, Qiang;Yang, Xinting;Sun, Chuanheng

作者机构：

关键词： Infrared imaging;Correlation model;3D trajectory

期刊名称：INFRARED PHYSICS & TECHNOLOGY （ 影响因子：2.638； 五年影响因子：2.581 ）

ISSN： 1350-4495

年卷期： 2018 年 88 卷

页码：

收录情况： SCI

摘要： Monitoring of fish-behavior has drawn extensive attention in pharmacological research, water environmental assessment, bio-inspired robot design and aquaculture. Given that an infrared sensor is low cost, no illumination limitation and electromagnetic interference, interest in its use in behavior monitoring has grown considerably, especially in 3D trajectory monitoring to quantify fish behavior on the basis of near infrared absorption of water. However, precise position of vertical dimension (z) remains a challenge, which greatly impacts on infrared tracking system accuracy. Hence, an intensity (L) and coordinate (z) correlation model was proposed to overcome the limitation. In the modelling process, two cameras (top view and side view) were employed synchronously to identify the 3D coordinate of each fish (x-y and z, respectively), and the major challenges were the distortion caused by the perspective effect and the refraction at water boundaries. Therefore, a coordinate correction formulation was designed firstly for the calibration. Then the L-z correlation model was established based on Lambert's absorption law and statistical data analysis, and the model was estimated through monitoring 3D trajectories of four fishes during the day and night. Finally, variations of individuals and limits of the depth detection of the model were discussed. Compared with previous studies, the favorable prediction performance of the model is achieved for 3D trajectory monitoring, which could provide some inspirations for fish behavior monitoring, especially for nocturnal behavior study. (C) 2017 Elsevier B.V. All rights reserved.

分类号：

  • 相关文献
作者其他论文 更多>>

微信小程序