基于深度学习的车位智能检测方法

Method for Intelligent Detection of Parking Spaces Based on Deep Learning

提出了一种基于深度学习的车位智能检测方法。利用TensorFlow深度学习平台对车辆目标识别模型进行了训练,提取了有效车辆图像的优化间隔,给出了车辆分布的精准识别结果,实现了对车辆分布识别结果的有序编号和车位空缺状况的准确判断。利用模拟数据和实际采集数据,分别验证了车位分布的智能识别、车位智能编号和空车位判断的可靠性。

Based on deep learning,one method for the intelligent detection of parking spaces is proposed.The TensorFlow deep learning platform is applied to train the car object recognition model,the optimal interval of the effective car images is extracted,the accurate recognition result of the car distribution is presented,and the order numbering of the recognition results of the car distribution and the accurate judgment of the vacancy situation of parking spaces are realized.The simulation results and the actually collected data are adopted to verify the reliability of intelligent identification of parking space distribution,intelligent numbering of parking space,and the judgement of empty parking space.