Analysis of the discriminative generalized Hough transform as a proposal generator for a deep network in automatic pedestrian and car detection
Authors:
- 7 June 2018
- journal article
- research article
- Published by SPIE-Intl Soc Optical Eng in Journal of Electronic Imaging
- Vol. 27 (05) , 051228-18
Abstract
Many approaches have been suggested for automatic pedestrian and car detection to cope with the large variability regarding size, occlusion, background variability, and aspect. Current deep learning-based frameworks rely either on a proposal generation mechanism (e.g., “faster R-CNN”) or on inspection of image quadrants/octants (e.g., “YOLO”), which are then further processed with deep convolutional neural networks (CNN). We analyze the discriminative generalized Hough transform (DGHT), which operates on edge images, for pedestrian and car detection. The analysis motivates one to use the DGHT as an efficient proposal generation mechanism, followed by a proposal (bounding box) refinement and proposal acceptance or rejection based on a deep CNN. We analyze in detail the different components of our pipeline. Due to the low false negative rate and the low number of candidates of the DGHT as well as the high accuracy of the CNN, we obtain competitive performance to the state of the art in pedestrian and car detection on the IAIR database with much less generated proposals than other proposal-generating algorithms, being outperformed only by YOLOv2 fine-tuned to IAIR cars. By evaluations on further databases (without retraining or adaptation), we show the generalization capability of our pipeline.Keywords
Funding Information
- Ministry of Education, Science and Cultural Affairs
This publication has 44 references indexed in Scilit:
- Fast Edge Detection Using Structured ForestsIEEE Transactions on Pattern Analysis and Machine Intelligence, 2014
- Fast Feature Pyramids for Object DetectionIEEE Transactions on Pattern Analysis and Machine Intelligence, 2014
- Strengthening the Effectiveness of Pedestrian Detection with Spatially Pooled Features2014
- Selective Search for Object RecognitionInternational Journal of Computer Vision, 2013
- IAIR-CarPed: A psychophysically annotated dataset with fine-grained and layered semantic labels for object recognitionPattern Recognition Letters, 2012
- Integral Channel Features2009
- Robust Object Detection with Interleaved Categorization and SegmentationInternational Journal of Computer Vision, 2007
- Object Detection Combining Recognition and Segmentation2007
- Detecting Pedestrians Using Patterns of Motion and AppearanceInternational Journal of Computer Vision, 2005
- Learning to detect objects in images via a sparse, part-based representationIEEE Transactions on Pattern Analysis and Machine Intelligence, 2004