Computer Based Detection of Alcohol Consumed Candidates Using Face Expressions with SIFT and Bag of Words

Nowadays, alcohol consumption is emerging as one of the most serious social issues. Many traffic accidents occur as a result of driving while inebriated. It poses a severe threat to the country, since many young people are losing their lives as a result of drunk driving on the highways, which also leads to traffic accidents. Totally, 80% of the road accidents are due to drunk driving conditions. Traffic police are conducting a drive to catch the drunk and drive cases with the Alcohol Breath Tester (ABT). ABT is effective in finding alcohol percentages. But here, there is a need of an automated system to analyze the alcohol consumpted candidate by recognizing their facial expression, this helps in identifying the alcohol consumed individuals very accurately. Here, a computer based detector is proposed by using human emotions. The proposed model with back propagation neural network leverages an accuracy of about 92%.