Pedestrian Detection in Crowded Environments through Bayesian Prediction of Sequential Probability Matrices

الملخص EN

In order to safely navigate populated environments, an autonomous vehicle mustbe able to detect human shapes using its sensory systems, so that it canproperly avoid a collision.

In this paper, we introduce a Bayesian approach tothe Viola-Jones algorithm, as a method to automatically detect pedestrians inimage sequences.

We present a probabilistic interpretation of the basicexecution of the original tool and develop a technique to produce approximateconvolutions of probability matrices with multiple local maxima.