Fast R-CNN Illegal parking enforcement system Using Fast R-CNN based on Vehicle detection

摘要

The present invention relates to an illegal parking/stoppage regulation system using a fast R-convolutional neural network (CNN) based vehicle detection capable of effectively performing illegal parking/stoppage regulation. According to one aspect of the present invention, the illegal parking/stoppage regulation system comprises: a first step of acquiring an image through a camera; a second step of dividing the entire area of an image into near and distant areas based on a reference line preset in the image; a third step of sectioning the image into a plurality of preset block units; a fourth step of using at least one of a plurality of convolution layers of a CNN to derive a feature map applied with the block units; a fifth step of using the feature map to determine whether an object included in the image is a vehicle; and a sixth step of recognizing a vehicle number from a license plate area of the vehicle when the object is a vehicle. When the object is included in the distant area, the feature map for an area including the object is derived by using the convolution layers. When the object is included in the near area, the feature map for an area including the object is derived by using only a predetermined number of layers among the convolution layers.