Search

Frequent analysis and updating the condition of traffic signs and mile markers are among the most important aspects of a highway asset management system. Today's practices mainly involve manual data collection and analysis, which have to be done for millions of miles of roads and the practice needs to be repeated regularly. While significant progress has been made on improving the data collection practice by leveraging video streams collected from car-mounted cameras, the analysis has primarily remained a manual and labor-intensive process. Automating the analysis from the collected videos is also challenging due to the interclass variability of traffic signs, expected changes in illumination, occlusion, sign position, and orientation. To address these challenges, this paper presents and evaluates the performance of three computer vision algorithms for detection and classification of traffic signs in presence of cluttered backgrounds and static and dynamic occlusions. The task particularly focuses on (1) extracting two-dimensional (2D) candidate windows from already collected video streams that potentially contain traffic signs-without making any prior assumption about their locations; (2) detecting the presence of signs in these 2D candidate windows; and (3) classifying them into warning, regulatory, stop, and yield sign categories based on their shape and color. For validation, a new comprehensive benchmark data set of over 11,000 annotated U.S. traffic sign images with a large range of pose, scale, background, illumination, and occlusion variation is introduced. Experimental results show an average accuracy of 76.20%, 89.31%, and 94.83% for the methods of (1) Haar-like features with Cascade classifiers, (2) histograms of oriented gradients (HOG) with multiple one-versus-all support vector machine (SVM) classifiers, and (3) HOG+C with the SVM classifiers, a variant of the second method with histograms of colors concatenated to HOG. The experimental results demonstrate the potential of leveraging joint representation of texture and color in HOG+C together with SVM discriminative classifiers as a viable solution for creating up-to-date and complete inventories of traffic signs for U.S. roadways. [ABSTRACT FROM AUTHOR]

A generalized traffic sign detection algorithm incorporating a hybrid active contour (HAC) model has been previously developed to automatically detect all types of traffic signs. Although the HAC model has shown some promising results, there are still some false negatives remaining due to the over-evolution of the HAC model. Therefore, further improvement is needed to reduce the number of these false negatives. This paper is aimed at developing a new geometry-preserving active polygon (GPAP) model to address the over-evolution issue in the HAC model for improving the computation speed. The contributions of this paper include (1) proposing a new geometry-preserving evolution that ensures that the "contour" only evolves at its vertices instead of at every point along the edges; (2) tailoring a new energy function that enables a polygon to effectively converge to traffic sign regions without losing its geometry integrity by using both local color contrast feature and global elongation and rectangularity features; (3) proposing a generalized Hough transform that can make an efficient initial guess for the active polygon based on the manual for uniform traffic control devices (MUTCD) sign shape template with a multiscale implementation. The experimental test shows that the proposed algorithm can achieve a per-sign detection rate of 85.6% out of 2,329 tested signs, while it can successfully avoid over-evolution and detect 233 traffic signs that could not be detected previously by the existing algorithm. By incorporating the proposed GPAP model into the existing generalized detection algorithm, the test on selected datasets with diverse image qualities, image contexts, and data acquisition configurations shows that the per-frame false negative rate is decreased by 12.2%; also, the average processing time reduces the existing algorithm's processing time by 50.2%. The proposed GPAP model shows the promise of improving both the accuracy and efficiency of the existing sign detection algorithm, which can be used to improve transportation agencies' image-based sign inventory. [ABSTRACT FROM AUTHOR]