基于VGGNet网络的雾霾无人车防碰撞研究

Collision Prevention Research of Haze Unmanned Vehicles Based on VGGNet Network

针对雾霾天气下无人车行驶容易出现视野受限,导致防碰撞能力下降的问题,提出了一种基于VGGNet网络的深度卷积神经网络模型,通过反向传播算法不断调整模型的权重和偏置,对收集雾霾天气下的图像和相关数据进行处理,实现模型的训练和优化。实验结果表明,所提出的方法可以有效地提高无人车在雾霾天气下的防碰撞能力,达到了良好的效果。研究结果可以为无人车行业在特殊气候条件下的防碰撞提供了一种新思路和实现方法,具有一定的参考价值和应用前景。

Aiming at the problem that unmanned vehicles driving under hazy weather are prone to have restricted vision,which leads to a decrease in collision avoidance ability,a deep convolutional neural network model based on VGGNet network is proposed,and the weights and biases of the model are continuously adjusted through the back-propagation algorithm,and the images and related data collected under hazy weather are processed to achieve the training and optimization of the model.The experimental results show that the proposed method can effectively improve the anti-collision ability of unmanned vehicles in hazy weather and achieve good results.The research results can provide a new idea and realization method for the unmanned vehicle industry to prevent collision under special weather conditions,which has certain reference value and application prospect.