Fuente: WIPO "processing tomato" The invention relates to the field of tomato planting decision making, in particular to a tomato planting decision making method based on a double-branch network and a computer device. The scheme comprises the following steps: acquiring and fusing multi-modal data; enhancing the fusion data through a self-attention generative adversarial network; feature extraction is carried out through a Transform-CNN (Convolutional Neural Network) dual-branch network; and reinforcement learning and decision making are carried out through DDPG. When the method is used for processing a complex structure and texture image, the self-attention layer can better grasp an image region relation, more accurate natural data are generated, a self-attention mechanism is fully utilized to capture global semantic information and a long-distance dependency relation, a CNN is utilized to extract local features and spatial information, effective fusion and feature extraction of multi-modal data are realized, and the method has the advantages of being high in robustness and high in robustness. And then a DDPG algorithm is used to solve a continuous action space problem, a deterministic strategy and an experience playback mechanism are combined, the accuracy of tomato planting decision making is greatly improved, and the tomato planting decision making method is suitable for tomato planting decision making.