基于目标深度分析的鱼体生长在线测量系统研究

Online Measurement System of Fish Growth Based on Target Depth Analysis

针对我国水产养殖中无法实时在线无接触地自动估测鱼体重量的难题,利用水下摄像头获取鱼体的双目图像,对水下图像进行优化,通过YOLACT实例图像分割模型对图像进行实时分割,通过自定义约束条件筛选的确信特征点完成左右图像匹配,最后计算出鱼体所处深度,从而复原鱼体实际大小。对鱼体在水下不同深度进行了测量准确性测试。结果表明,在水深70 cm处存在测量最优值,长度和宽度误差分别为6.50%和5.75%。本研究为实现水产养殖中的个体实时无接触生长监测,从而形成闭环的物联网化生产模型提供了一种可行方案。

Aiming at the problem that it is impossible to automatically estimate the weight of fish in real-time,online and contactless in China aquaculture,we used the underwater camera to obtain the binocular image of the fish body and optimized the underwater image.Then,we segmented the image in real-time through YOLACT instance image segmentation model,completed the left and right image matching by the confident feature points filtered by user-defined constraints,calculated the depth of the fish,and restored the actual size of the fish.Measurement accuracy tests were conducted on the fish body at different depths underwater.The results showed that there was an optimal measurement value at a depth of 70 cm,and the length and width errors were 6.50%and 5.75%,respectively.This study provides a feasible scheme to realize real-time contactless growth monitoring of individuals in aquaculture and form a closed-loop internet of things production model.