基于图像处理的鱼卵计数方法研究

针对传统的鱼卵计数方法效率低、精度差、对鱼卵有损伤等缺点，提出了基于图像处理的鱼卵计数方法: 通

过Matlab 软件中强大的图像处理功能，对经过Photoshop 预处理的图像进行灰度化、去除噪音、二值化和形态学处

理，采用连通图计数和面积计数法进行鱼卵的计数，精确性在95 以上。该方法减轻了操作者劳动强度，弥补了

人视觉的不足之处，可以有效地提高工作效率和计数精度，并且可避免鱼卵损伤，有利于后续工作的开展，具有一

定的理论和应用价值。-In view of the traditional method of counting fish spawns has many shortcomings，such as poor precision，

low efficiency and eggs damage． In order to cover the shortcomings，being helpful for follow-up work，two methods

of counting fish spawns based on image processing was proposed in this paper． Through the powerful image processing

function of Matlab software，putting up median filter，gray processing，binary processing and morphological operation

to the reading image dealt by Photoshop software． Then on the theory of connected graph counting and area

comparison，counting up the number of spawns，the degree of counting precision is higher than 95　which have

certain theoretical and practical value．