Name
Abstract | ||
|---|---|---|
Complex weather factors seriously affect the quality and accuracy of image processing, and indirectly pose severe challenges to the expansion of IoT applications in smart cities and the actual intelligent classification of research. However, the current research on weather classification cannot comprehensively deal with composite weather classification, and the accuracy and processing time are still not satisfactory. This is still a challenge for the popularization of IoT applications. This paper proposes a multi-weather classification method that can intelligently classify the weather elements in the image data collected by the IoT applications. The idea behind us is to propose the evolutionary algorithm into EfficientNet to solve the various weather classification problem. Our experimental results show that we can achieve remarkable classification results with a high degree of confidence in five complex weather conditions, even in pictures with obvious weather characteristics, and the recognition accuracy increases to more than 95%. |
Year | DOI | Venue |
|---|---|---|
2021 | 10.1109/PERCOMWORKSHOPS51409.2021.9430939 | 2021 IEEE INTERNATIONAL CONFERENCE ON PERVASIVE COMPUTING AND COMMUNICATIONS WORKSHOPS AND OTHER AFFILIATED EVENTS (PERCOM WORKSHOPS) |
Keywords | DocType | Citations |
Multi-Weather Classification, Image Processing, Evolutionary Algorithm, EfficientNet, IoT | Conference | 0 |
PageRank | References | Authors |
0.34 | 0 | 6 |
Name | Order | Citations | PageRank |
|---|---|---|---|
| Yulong Zhang | 1 | 0 | 0.34 |
| Jingtao Sun | 2 | 4 | 4.85 |
| Mingkang Chen | 3 | 0 | 0.34 |
| Qiang Wang | 4 | 0 | 3.04 |
| Yuan Yuan | 5 | 301 | 26.63 |
| Rongzhe Ma | 6 | 0 | 0.34 |