Craig Henderson

PhD Research Student exploring techniques in Computer Vision, Machine Learning using C++11

My research interests centre around object tracking and retrieval in video and temporal still images, particularly in low frame-rate and multi-angle scene imaging such as CCTV. I am experimenting with object search and re-identification with a goal to locating offenders in previous footage by way of distinctive clothing or markings such as tatoos.

Older projects, publications and research can be found here

Publications

Large-scale forensic analysis of security images and videos

with Professor Ebroul Izquierdo

6th Doctoral Consortium, BMVC 2014, Nottingham, 5th September 2014



Our research is concerned with the practical application of computer vision in the forensic analysis of security images and videos. Contemporary literature make use of high-definition images and Hollywood feature films in their datasets, and there is little or no assessment of algorithms' performance using poor quality images with variable frame rates and uncontrolled lighting conditions such as security video.

Work so far has produced a methodology for matching features across low quality images that yields an improved results over existing feature matching techniques. Future work will involve innovation in search and retrieval online machine learning to train models from unlabelled data, and segmentation of one-shot videos to aid computer and human analysis of long-running video sequences. We are motivated to produce an integrated system for police investigators to use a query-by-example search and retrieval system with relevance feedback and machine learning to incrementally discover evidence in criminal investigations.

@inproceedings{Henderson2014,
address = {Nottingham},
author = {Henderson, Craig and Izquierdo, Ebroul},
booktitle = {6th Doctoral Consortium at BMVC 2014},
title = {{Large-scale forensic analysis of security images and videos}},
year = {2014}
}
                            

Boosting feature correspondence in low quality security camera videos using color information

with Professor Ebroul Izquierdo

under review, 15th August 2014



Contemporary feature extractors focus on invariance to scale, rotation and illumination changes, and typically produce a descriptor that encodes the texture of the image part close to the detected feature. In processing low quality security video images, the poor image definition, blurring and noise between frames can cause a much reduced accuracy in finding similar features. In this paper, we describe a method of boosting the matching performance of standard feature descriptors by means of a small extension to encode color information from the image pixels surrounding the feature key point or centre point, and a mathematically proven distance calculation between two descriptors.

We test our method on a variety of real-world security videos and demonstrate an improvement in the ability to match features between images compared with the naked feature descriptors extracted from the same set of feature points. We use key point and region based features from Harris Corners, BRISK, SIFT and SURF as well as MSER and MSCR to provide a comprehensive analysis of our generic extension method. Our results show an improvement of up to 85% using BRISK features and up to 144% improvement by applying our method to the RootSIFT descriptor using Euclidean distance.

Boosting feature matching accuracy with colour information PDF (poster)

with Professor Ebroul Izquierdo

BMVA Summer School, Swansea, May 2014

ViiHM Workshop, Stratford-upon-Avon, September 2014



We boost the performance of discriminative matching of features between colour images, with

  • a space-efficient and generic extension for any feature descriptor
  • a novel distance measure calculation