Open AccessResearch and application of small target detection method
Author(s) -
Zhenheng Wang,
Huajun Wang,
Jun Wan,
Yu Ma,
Chuanfeng Huang,
Yongyi Yang
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2029/1/012129
Subject(s) - computer science , artificial intelligence , image (mathematics) , object detection , train , computer vision , pattern recognition (psychology) , data mining , cartography , geography
With the advent of the big data era, more and more data are used in image detection. In order to solve the problem of small target detection in traditional image detection algorithms, this paper proposes an improved target detection algorithm based on YOLO-v5. The algorithm proposes to connect two models in series, using the former model for target pre-positioning, and the latter model for target detection. This method trains two different model parameters, tests them separately, and finally connects them for target detection. The test results show that the improved algorithm improves the ability in small target detection, and at the same time has a speed advantage.