Scalable non-parametric parsing for segmentation and recognition of high-quantity, low-cost highway assets from car-mounted video streams

Keyphrases

• Parsing 100%
• Nonparametric 100%
• Video Stream 100%
• Highway Assets 100%
• Superpixel 75%
• Graph-based Segmentation 50%
• State-of-the-art Techniques 25%
• Histogram 25%
• Further Development 25%
• Training Set 25%
• Texton 25%
• Likelihood Ratio 25%
• Segmentation Algorithm 25%
• Markov Random Field 25%
• Video Frames 25%
• Ratio Measures 25%
• Virginia Tech 25%
• Smart Roads 25%
• SIFT Descriptor 25%
• 2D Video 25%
• Parsing Algorithm 25%
• Condition Assessment 25%
• Markov Random Field Optimization 25%
• Dense SIFT 25%
• Video-based Assessment 25%
• Global Matching 25%
• 3D Assets 25%
• Traffic Signs 25%
• Guardrail 25%
• Light Pole 25%

Engineering

• Random Field 100%
• Limitations 50%
• Experimental Result 50%
• Histogram 50%
• Likelihood Ratio 50%
• State-of-the-Art Method 50%
• Road 50%
• Virginia Tech 50%
• Condition Assessment 50%

Computer Science

• Parsing 100%
• Superpixel 100%
• Experimental Result 33%
• Likelihood Ratio 33%
• segmentation algorithm 33%
• Scale-Invariant Feature Transform 33%