Efficient Pedestrian Detection in Top-View Fisheye Images Using Compositions of Perspective View Patches

TL;DR

This paper introduces a method for pedestrian detection in top-view fisheye images by creating composite perspective views, enabling the use of existing detectors without retraining, and mapping detections back to fisheye frames.

Contribution

The paper proposes a novel approach to generate composite perspective views from fisheye images, allowing existing detectors to be used directly without retraining, and introduces a new mapping method for bounding boxes.

Findings

Detection performance compares favorably with state-of-the-art methods.

The approach effectively handles orientation variation in fisheye images.

Existing perspective detectors can be applied directly to composite images.

Abstract

Pedestrian detection in images is a topic that has been studied extensively, but existing detectors designed for perspective images do not perform as successfully on images taken with top-view fisheye cameras, mainly due to the orientation variation of people in such images. In our proposed approach, several perspective views are generated from a fisheye image and then concatenated to form a composite image. As pedestrians in this composite image are more likely to be upright, existing detectors designed and trained for perspective images can be applied directly without additional training. We also describe a new method of mapping detection bounding boxes from the perspective views to the fisheye frame. The detection performance on several public datasets compare favorably with state-of-the-art results.