Real time traffic sign recognition
First Claim
Patent Images
1. A method for recognition of traffic signs, the method comprising:
- receiving an input image of a captured scene;
identifying one or more approximate locations of interest in the input image based at least partially on extracting color information from the input image, wherein extracting the color information includes extracting initial binary masks from the input image that include a red initial binary mask, a blue initial binary mask, a yellow initial binary mask, and a white initial binary mask, and applying, to each of the red, blue, yellow, and white initial binary masks, a morphological opening operation that includes an erosion operation follows by a dilation operation to produce respective red, blue, yellow, and white final binary masks;
processing each of the one or more approximate locations of interest in the input image to identify one or more localized windows in the input image in which a traffic sign is present;
for each of the one or more localized windows, determining initial classification information of the respective traffic sign present in the localized window based on a plurality of traffic sign classification categories;
applying temporal smoothing to each of the one or more localized windows to determine, for each of the one or more localized windows, final classification information of the respective traffic sign present in the localized window; and
outputting information that identifies, for each traffic sign present in the captured scene, a traffic sign type based on the final classification information.
1 Assignment
0 Petitions
Accused Products
Abstract
Advanced driver assistance systems need to be able to recognize and to classify traffic signs under real time constraints, and under a wide variety of visual conditions. The invention shown employs binary masks extracted by color space segmentation, with a different binary mask generated for each sign shape. Temporal tracking is employed to add robustness to the detection system. The system is generic, and is trainable to the signs used in various countries.
8 Citations
17 Claims
-
1. A method for recognition of traffic signs, the method comprising:
-
receiving an input image of a captured scene; identifying one or more approximate locations of interest in the input image based at least partially on extracting color information from the input image, wherein extracting the color information includes extracting initial binary masks from the input image that include a red initial binary mask, a blue initial binary mask, a yellow initial binary mask, and a white initial binary mask, and applying, to each of the red, blue, yellow, and white initial binary masks, a morphological opening operation that includes an erosion operation follows by a dilation operation to produce respective red, blue, yellow, and white final binary masks; processing each of the one or more approximate locations of interest in the input image to identify one or more localized windows in the input image in which a traffic sign is present; for each of the one or more localized windows, determining initial classification information of the respective traffic sign present in the localized window based on a plurality of traffic sign classification categories; applying temporal smoothing to each of the one or more localized windows to determine, for each of the one or more localized windows, final classification information of the respective traffic sign present in the localized window; and outputting information that identifies, for each traffic sign present in the captured scene, a traffic sign type based on the final classification information. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
Specification