鱼目标三维信息采集平台中干扰现象的判定与消除

为了保证鱼类行为数据采集的准确性和鲁棒性,本实验提出了一种干扰消除算法,对鱼目标三维定位算法进行优化。主要针对鱼目标三维信息采集平台中出现的一种由全反射引起的干扰现象,利用光学原理分析其出现原因,总结干扰现象特点,之后通过基于轮廓的运动目标跟踪方法,设计干扰消除算法,优化鱼目标三维定位算法,实现对干扰现象影响的消除,提高定位精度。通过对红鲫和仿真鱼的定位实验,结果发现,本实验提出的去干扰优化算法可以有效消除干扰像对于目标定位的影响,准确获取鱼目标的三维坐标,保证鱼目标三维定位跟踪的准确性和鲁棒性,提高鱼目标行为信息采集精度。

Interference phenomenon in a 3D information acquisition platform of fish target

In order to ensure the accuracy and robustness of fish behavior data acquisition, an interference cancellation algorithm was proposed to optimize the 3D positioning algorithm of fish target.In this paper, the study was based on a singular phenomen-on caused by total reflection in this platform, the causes of the phenomenon were analyzed, and the influence of the phenomenon was eliminated. Firstly, based on the principle of geometrical optics and the principle of total reflection, the causes of interference phenomena were analyzed, and the characteristics of interference image were summarized to determine the elimination scheme.Using the digital image and video processing technology, the interference recognition and elimination algorithm was designed to optimize the original positioning algorithm.Based on the experiment of red crucian carp and simulated fish, the results showed that the proposed algorithm is accurate and confirmed that this algorithm has high precision and robustness.