The invention discloses a road asset detection model construction method and a road asset detection method. The method comprises the following steps: acquiring a road asset image data set; classifying images in the road asset image data set according to target attributes, and inputting a classified image sample set with labels into a preset neural network for training until a road asset detection model meeting conditions is obtained; and performing frame extraction processing on the acquired video stream data to obtain a corresponding road asset image data set, and inputting the road asset image data set into a road asset detection model for training. According to the method, images in an image data set are classified according to target attributes, so that inference of attributes of different labels by a model is facilitated, management and deployment of the model are facilitated, and road asset images can be comprehensively and accurately mastered; and then the trained road asset detection model is utilized to detect the input road asset image data, manual detection is replaced, and the detection efficiency and accuracy are improved.
本发明公开了一种道路资产检测模型构建和道路资产检测方法,获取道路资产图像数据集;对所述道路资产图像数据集中的图像按照目标属性进行分类并将分类后带标签的图像样本集输入至预设的神经网络中训练直至得到满足条件的道路资产检测模型;对获取的视频流数据进行抽帧处理得到对应的道路资产图像数据集并输入道路资产检测模型进行训练。该方法对图像数据集中的图像按照目标属性进行分类,有助于模型对于不同标签的属性的推断,同时方便了模型的管理和部署可以全面、准确地掌握道路资产图像;然后利用训练好的道路资产检测模型对输入的道路资产图像数据进行检测,代替了人工检测,提高了检测效率和准确性。
Road asset detection model construction and road asset detection method
一种道路资产检测模型构建和道路资产检测方法
2022-06-24
Patent
Elektronische Ressource
Chinesisch
Managing the Road Asset: Managing the Road Asset in Australia and New-Zealand
British Library Online Contents | 1997
New Zealand road asset management
British Library Conference Proceedings | 1994
|Road facility evaluation-based road traffic asset rationality analysis method
Europäisches Patentamt | 2024
|Asset management for sustainable road funding
TIBKAT | 2013
|Road trains - a positive asset for road park & ride schemes
British Library Online Contents | 1994