计算机视觉技术的车型识别应用研究

针对目前立体停车库主要采用传统的传感器入库检测系统来识别车辆信息,存在施工周期长、器件损坏率较高,维护成本较高的问题,提出了一种应用计算机视觉技术的车型识别解决方案.应用卷积神经网络框架Caffe,基于Caf-feNet模型,通过fine-tuning模型优化以及参数优化,最终得到了一个性能较优异的识别模型.实验结果表明,该模型可克服输入图片背景复杂多变,目标被遮挡的情况,对轿车车型识别这一问题鲁棒性好,具有一定的可行性及应用价值.

Application Research on Vehicle Recognition Based on Computer Vision Technology

The current stereo garage equipment mainly adopts the traditional sensor detection system to identify the vehicle information, but exists the problems such as long construction period, high damage rate and high maintenance cost.So a model is proposed, which applies computer vision technology to solve the problems of Vehicle recognition.Applying Caffe convolution neural network framework, proceeding the fine-tuning model optimization and parameter optimization based on CaffeNet model, finally a recognition excellent model is achieved.The experiment results show that the model can handle the input image with complex background and the target obscured.The model has a good robustness to identify the type of the car and certain feasibility and application value.