Abstract
Target detection is the basic technology of self-driving system. In this paper, the problem of high detection rate of pedestrians and other small targets is studied in real-time detection of Tiny YOLOV3 target detection algorithm, and the network structure of Tiny YOLOV3 algorithm is improved. 2-step convolutional layers are added to the network, and deep separable convolution constructs are used to replace the traditional convolutions. On the basis of the original two-scales prediction target of the network, a scale is added to form a three-scales prediction, which can makes the detection of small targets such as pedestrians more accurate. The experimental results show that the average accuracy of the improved target detection algorithm is 8.6% higher than that of Tiny YOLOV3, and it meets the real-time requirements and has certain robustness.
Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callMore From: IOP Conference Series: Earth and Environmental Science
Disclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
