基于卷积神经网络的道路多目标检测方法 |
| |
引用本文: | 迟志诚.基于卷积神经网络的道路多目标检测方法[J].汽车实用技术,2021(7):23-24. |
| |
作者姓名: | 迟志诚 |
| |
作者单位: | 长安大学汽车学院 |
| |
摘 要: | 得益于数字图像处理技术快速的发展和计算机硬件性能的提高,基于机器学习和深度学习的图像处理技术,成为智能驾驶视觉感知的重要支撑。为了在实际道路环境中持续高效的检测道路目标,文章利用了YOLO神经网络作为主要检测框架。使用卷积神经网络可以同时捕捉到目标的底层和高层特征。物体的底层特征可以符合人的视觉感知特征和主观感受,确定物体的所属种类和外观形状,将底层特征与高层语义特征结合进一步增强神经网络识别的准确度和鲁棒性。
|
关 键 词: | 卷积神经网络 目标识别 自动驾驶 YOLO v3 |
Multiple Object Detection Method Based on Convolutional Neural Networks on Road |
| |
Authors: | Chi Zhicheng |
| |
Institution: | (Chang'an University,School of Automobile,Shaanxi Xi’an 710064) |
| |
Abstract: | Because of the rapid development of digital image processing technology,image processing technology based on machine learning has become an important support for the visual perception of intelligent vehicle.In order to continuously and efficiently detect targets in the actual road environment,we use the YOLO v3 neural networks as the main detection framework.This network can simultaneously capture the bottom and high-level features of the target.The low-level features of an object can conform to human visual perception features and subjective feelings,determine the type and appearance of the object.We combined the low-level features with high-level semantic features to further enhance the accuracy and robustness of neural network recognition. |
| |
Keywords: | Convolutional neural networks Object recognition Autonomous driving YOLO v3 |
本文献已被 维普 等数据库收录! |
|