基于BP神经网络的羊肚菌液体发酵量化方法

Quantified Method for Morchella Submerged Fermentation Based on BP Neural Network

根据羊肚菌液体种子在三角瓶内生长呈球体的特点,可推知液体种子在三角瓶底部的图像映射面积对应于液体种子的质量,利用图像处理和人工神经网络可以实时并准确得到不同时刻的液体种子质量,从而绘制出羊肚菌液体种子的生长曲线,进一步确定恰当的接种时机。本文建立了羊肚菌液体发酵确定接种菌龄的量化方法,经验证,此方法避免了人为误差,结果比较准确。

The growth curve of liquid seeds were drawed by the way of computer image treatment and ANN after studying the shape characteristic of Morchella. The quantified method of judging inoculating age of Morchella were established, which has avoided human error and the result was satisfactory.