HOG and Gabor Filter Based Pedestrian Detection using Convolutional Neural Networks

Pedestrian detection is an essential research topic due to its major importance especially in the fields of automotive, surveillance and robotics. In spite of having tremendous improvements, pedestrian detection is still an open challenge and searching for more and more accurate algorithms. In the last few years, deep learning more specifically Convolutional Neural Networks emerged as the state of the art in terms of accuracy for a number of computer vision tasks such as image classification, object detection, segmentation, etc. In this paper, we have proposed a pedestrian detection system based on the Gabor filter, Histogram of Oriented Gradient and Convolutional Neural Networks. two mostly used datasets, INRIAPerson and Daimler Mono Pedestrian dataset for pedestrian detection is used for both training and testing. The PennFidanPed dataset is used for testing only. From experimental results, it is shown that we have accomplished comparatively better accuracy close to state of the art approaches.