Driving state recognition method and system based on improved bilinear CNN model

• Main Authors: ZHOU HUI, WU YIZHUO, ZHU XIAOHAO, YANG DINGLI
• Format: Patent
• Language: Chinese; English
• Published: 19.09.2023
• Subjects: CALCULATING; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS; COMPUTING; COUNTING; PHYSICS

The invention discloses a driving state identification method and system based on an improved bilinear CNN model. The method comprises the following steps: S1, respectively carrying out LALBP feature extraction and histogram equalization on a to-be-processed image sequence; the to-be-processed image sequence is a collected expression image sequence of a driver driving a vehicle; s2, inputting the image after the LALBP feature extraction into an improved bilinear CNN model to obtain a first image feature; the method comprises the following steps: S1, carrying out histogram equalization on an image to obtain a first image feature, S2, carrying out histogram equalization on the image, S3, inputting the image subjected to histogram equalization into an improved bilinear CNN model to obtain a second image feature, S4, collecting the first image feature and the second image feature into a bilinear vector through bilinear operation, and finally, outputting a classification probability and classification through a fu