| CPC G06V 20/54 (2022.01) [G06N 3/08 (2013.01); G06T 7/60 (2013.01); G06T 7/73 (2017.01); G06T 7/90 (2017.01); G06V 10/56 (2022.01); G06V 10/82 (2022.01); G06T 2207/10024 (2013.01); G06T 2207/20081 (2013.01); G06T 2207/20084 (2013.01); G06T 2207/30236 (2013.01); G06V 2201/07 (2022.01)] | 15 Claims |
|
1. A method for evaluating environment of a pedestrian passageway comprising:
obtaining a position information of a target area in an environment of a pedestrian passageway;
obtaining a streetscape image corresponding to the position information of the target area, wherein the streetscape image comprises a plurality of target objects;
inputting the streetscape image into a trained convolutional neural network, making the trained convolutional neural network carry out a convolution calculation of the streetscape image to generate a feature vector for classifying the plurality of the target objects in the streetscape image, and outputting the feature vector, comprising: using the trained convolutional neural network to carry out a convolution calculation of the streetscape image to generate a feature map of the streetscape image; using an object recognition and segmentation model to take each point of the feature map of the streetscape image as center points of frames; comparing each of the frames with the streetscape image to determine each of the target objects in the streetscape image; outputting a plurality of target frames that framing each of the target objects in the streetscape image, and classifying each of the target objects to obtain the feature vector;
inputting the feature vector into a full convolution neural network to label a plurality of pixels belonging to a same target object by color, and outputting the streetscape image labeling a plurality of target objects with color.
|