The invention discloses a
cibotium barometz
sporophyte transformation method and
system based on multi-factor
coupling regulation, and relates to the technical field of
biology. The method comprises the following steps: synchronously acquiring environment data and
sporophyte image data; preprocessing the environment data, performing
feature extraction on the image data, and splicing the processed data into a comprehensive
feature vector; quantifying the weight, correlation degree and interaction level among the environmental factors through an entropy method,
grey correlation analysis and a
coupling coordination degree model; constructing an improved CNN-LSTM neural network, extracting features through a double-
branch structure, fusing a
coupling analysis result to carry out cross-
modal fusion, and training to obtain a prediction model; when the predicted conversion rate is lower than a threshold value, environment parameters are automatically adjusted until conversion is completed; according to the method, the problems of
environmental factor data isolation and regulation
lag in the prior art are solved, accurate prediction and intelligent regulation of the
sporophyte transformation process are realized, and the transformation success rate and the production efficiency are remarkably improved.