Low-illumination image enhancement method based on NASPE-net

The invention discloses a low-illumination image enhancement method based on NASPE-net. The method comprises the following steps: 1) dividing a data set into a low-illumination image and a normal-illumination image according to the illumination degree; 2) inputting the low-illumination image into NASPE-net to extract a parameter mapping graph; the invention discloses a low-illumination image enhancement method, which comprises the steps of (1) obtaining a low-illumination image, (2) obtaining a low-illumination image, (3) generating an attention weight map of an RGB channel, (4) obtaining an enhanced image similar to a normal-illumination image, (5) adopting a GAN method to resist loss, and (6) designing a loss function to reversely train an NASPE-net network.The low-illumination image enhancement method does not need paired low/normal-brightness image data sets, is small in parameter quantity, and can well enhance a low-illumination image. 本发明公开一种基于NASPE-net的低照度图像增强方法，包括如下步骤：1）对数据集依照光照程度划分为低照度图像和正常照度图像；2）将低照