Lung parenchyma image segmentation method based on Unet model

The invention provides a lung parenchyma image segmentation method based on a Unet model, solves the problems that a lung parenchyma segmentation model based on deep learning is high in complexity, large in calculation amount and multiple in parameter, and belongs to the field of medical image processing. The method comprises the following steps: acquiring a lung parenchyma CT image to be segmented; preprocessing the image; establishing an LUnet network structure: on the basis of an original Unet model, replacing an original convolution module in a coding path with a pre-activated residual module, introducing a multi-scale context module between the coding path and a decoding path, reducing the number of network layers and adjusting the number of channels of each layer; training the LUnet network by using the preprocessed pulmonary parenchyma CT image; and inputting a lung parenchyma CT image test set to be segmented into the trained LUnet network to obtain a segmentation result. Experimental results show tha