New energy heavy truck back type battery pack detection method and device based on deep learning and readable medium

• Main Authors: WU YONGXIONG, CHEN WEI, GUAN RONG, CHU XUYANG, ZHANG JINYI, PANG MINGLONG, CHEN XIAOMENG, SONG XIANGYANG, HU ZHANPENG, LIN WEILONG, XIANG RUIHENG
• Format: Patent
• Language: Chinese; English
• Published: 03.05.2024
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; IMAGE DATA PROCESSING OR GENERATION, IN GENERAL; PHYSICS

The invention discloses a new energy heavy truck back type battery pack detection method and device based on deep learning and a readable medium. The new energy heavy truck back type battery pack detection method comprises the steps that a to-be-detected heavy truck back type battery pack top image is acquired; constructing a heavy truck back type battery pack detection model based on an improved YOLOv5 network and training the heavy truck back type battery pack detection model to obtain a trained heavy truck back type battery pack detection model; the heavy truck back type battery pack detection model is a network structure formed by adding a double-layer routing attention module between a first convolutional layer and a first up-sampling layer of a head structure in a standard YOLOv5 network model; and inputting the top image of the back type battery pack of the heavy truck into the trained detection model of the back type battery pack of the heavy truck, and identifying the position and category informatio