基于计算机视觉的鱼体色明暗程度量化方法

当前．鱼类行为学在养殖生产中的应用还不普遍，但已经引起人们的重视。借助于现代技术手段监涸鱼的行为，研究不同应激条件下鱼的行为特点以及鱼的行为与水产养殖环境的关系。从而实现基于鱼行为变化的应激自动报警。是很有意义的。为此。利用计算机视觉技术近似量化水中鱼体色的明暗程度。将图像转换到HSV空间得到鱼体的明度值，与经校准后的无色玻璃覆盖的灰级各阶的明度值进行比较和线性擂值。近似用明度阶值表示鱼体色的亮暗程度。

Quantifying the Fish Skin Darkness Using Computer Vision

Skin darkening in some fish appears to be related to stress and illness. The fish skin darkness can be measured continuously in-situ and has the potential to be used in the actual productions to predict stress conditions and healthy states. This paper presents a method of quantifying the skin darkness of the fish swimming in water using computer vision technique. After converting the image from RBG color space to HSV color space, the lightness of the fish body was calculated and compared with the lightness of density steps of the gray scale covered with transparent glass calibrated by the gray scale, the skin darkness was denoted by density steps of the linear black-white scale.