Aiming at the problems of low real-time and low accuracy in target recognition of pedestrians and vehicles in infrared images, as well as missing long-distance detection, a YOLO-based deep learning method for pedestrian and vehicle recognition was proposed. First, image preprocessing and data enhancement are applied to improve the generalization ability of the model; then the network structure were changed, as well as the multi-scale feature detection map and the PANet structure were applied to improve the recognition ability of different scale targets; finally the attention mechanism was applied to improve the detection accuracy. In the simulation, the actual collected infrared pedestrian and vehicle data sets were used for training, and got the training model. Then the proposed method was transplanted to the embedded GPU platform to verify the performance of the algorithm. The result shows that the recognition accuracy rate reaches 84.35%, 12.44% improved compared with the original YOLO method, and the speed has reached 88 frames per second, which can meet the actual engineering needs.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.