Road thrown object detection method based on contrast clustering self-learning

The invention discloses a road thrown object detection method based on contrast clustering self-learning, and the method comprises the steps: constructing a road thrown object detection network model which comprises a feature extraction module, a candidate region module, an unknown label generation module and a thrown object recognition module based on a maximum logic value. Training the road throwing object detection network model by adopting a training sample, detecting an image to be detected by adopting the trained road throwing object detection network model, and outputting a detection result. According to the method, the thrown objects static on the expressway can be recognized, the anti-interference capability is high, thrown object targets not included in the thrown object types in the data set can be detected, and therefore the method has the practical value. 本发明公开了一种基于对比聚类自学习的道路抛洒物检测方法，构建道路抛洒物检测网络模型，所述道路抛洒物检测网络模型包括特征提取模块、候选区域模块、未知标签生成模块和基于最大逻辑值的抛洒物识别模块。采用训练样本训练所述道路抛洒物检测网络模型，采用训练好的道路抛洒物检测网络模型对待检测图像进行