基于上下文信息的监控场景行人检测方法

Pedestrian detection method in surveillance scenesbased on context information

为了提高监控场景中行人检测的准确度,提出了一种基于上下文信息的行人检测方法。该方法将监控场景的上下文信息融入到卷积神经网络中,选择性地学习对行人检测有帮助的上下文信息。首先,利用一个截断的卷积神经网络提取输入图像的多张特征图。然后,将多张特征图通过两个包含上下文信息的卷积层,形成一张掩码图。最后,通过在掩码图上估计行人的边界框,获得行人检测的结果。实验表明,该方法能实现监控场景中准确且快速的行人检测。

In order to improve the accuracy of pedestrian detection in surveillance scenes, a pedestrian detection method based on context information is proposed. This method combines the context information of surveillance scenes into a convolutional neural network, which can optionally learn the context information that are helpful for pedestrian detection. Firstly, a truncated convolutional neural network is used for extractingmultiple feature maps according to the input image. Then, these feature maps pass through two contextual convolutional layers to form a mask. Lastly, the pedestrian detection results are obtained through estimating the bounding boxes on the mask. The experiments show that our method can implement the precise and fast pedestrian detection in surveillance scenes.