Cabin door low-quality image optimization method based on deep learning

• Main Authors: ZUO XIAOBIN, LI XU, CHEN MUHUA, SHI JIANPING
• Format: Patent
• Language: Chinese; English
• Published: 28.04.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention discloses a cabin door low-quality image optimization method based on deep learning, and the method comprises the following steps: collecting a cabin door docking image, and making a high-quality-low-quality image data set; the method comprises the following steps: constructing a deep learning network of a three-stage architecture based on a U-net architecture encoder-decoder and a single-scale channel structure; introducing an attention mechanism between stages of the deep learning network; image refining modules are respectively introduced between the encoder and the decoder and between the encoder and the decoder and the single-scale channel structure; and training the deep learning network. According to the method, based on deep learning, the low-quality image acquired in the cabin door docking process is optimized, so that the image features are clearer, and the whole docking process of the cabin door and the gallery bridge is smoother and more accurate. 一种基于深度学习的舱门低质量图像优化方法，包括以下步骤：采集舱门对接图像