Histogram of Oriented Gradients with Cell Average Brightness for Human Detection
A modification of the descriptor in a human detector using Histogram of Oriented Gradients (HOG) and Support Vector Machine (SVM) is presented. The proposed modification requires inserting the values of average cell brightness resulting in the increase of the descriptor length from 3780 to 3908 valu...
Main Author: | |
---|---|
Format: | Article |
Language: | English |
Published: |
Polish Academy of Sciences
2016-03-01
|
Series: | Metrology and Measurement Systems |
Subjects: | |
Online Access: | http://www.degruyter.com/view/j/mms.2016.23.issue-1/mms-2016-0012/mms-2016-0012.xml?format=INT |
id |
doaj-9e35e4bc0f81489ba96a78fe51714f35 |
---|---|
record_format |
Article |
spelling |
doaj-9e35e4bc0f81489ba96a78fe51714f352020-11-25T02:48:08ZengPolish Academy of SciencesMetrology and Measurement Systems2300-19412016-03-01231273610.1515/mms-2016-0012mms-2016-0012Histogram of Oriented Gradients with Cell Average Brightness for Human DetectionWójcikowski Marek0Gdańsk University of Technology, Faculty of Electronics, Telecommunications and Informatics, G. Narutowicza 11/12, 80-233 Gdańsk, PolandA modification of the descriptor in a human detector using Histogram of Oriented Gradients (HOG) and Support Vector Machine (SVM) is presented. The proposed modification requires inserting the values of average cell brightness resulting in the increase of the descriptor length from 3780 to 3908 values, but it is easy to compute and instantly gives ≈ 25% improvement of the miss rate at 10‒4 False Positives Per Window (FPPW). The modification has been tested on two versions of HOG-based descriptors: the classic Dalal-Triggs and the modified one, where, instead of spatial Gaussian masks for blocks, an additional central cell has been used. The proposed modification is suitable for hardware implementations of HOG-based detectors, enabling an increase of the detection accuracy or resignation from the use of some hardware-unfriendly operations, such as a spatial Gaussian mask. The results of testing its influence on the brightness changes of test images are also presented. The descriptor may be used in sensor networks equipped with hardware acceleration of image processing to detect humans in the images.http://www.degruyter.com/view/j/mms.2016.23.issue-1/mms-2016-0012/mms-2016-0012.xml?format=INTdigital image processingobject detectionhuman detection |
collection |
DOAJ |
language |
English |
format |
Article |
sources |
DOAJ |
author |
Wójcikowski Marek |
spellingShingle |
Wójcikowski Marek Histogram of Oriented Gradients with Cell Average Brightness for Human Detection Metrology and Measurement Systems digital image processing object detection human detection |
author_facet |
Wójcikowski Marek |
author_sort |
Wójcikowski Marek |
title |
Histogram of Oriented Gradients with Cell Average Brightness for Human Detection |
title_short |
Histogram of Oriented Gradients with Cell Average Brightness for Human Detection |
title_full |
Histogram of Oriented Gradients with Cell Average Brightness for Human Detection |
title_fullStr |
Histogram of Oriented Gradients with Cell Average Brightness for Human Detection |
title_full_unstemmed |
Histogram of Oriented Gradients with Cell Average Brightness for Human Detection |
title_sort |
histogram of oriented gradients with cell average brightness for human detection |
publisher |
Polish Academy of Sciences |
series |
Metrology and Measurement Systems |
issn |
2300-1941 |
publishDate |
2016-03-01 |
description |
A modification of the descriptor in a human detector using Histogram of Oriented Gradients (HOG) and Support Vector Machine (SVM) is presented. The proposed modification requires inserting the values of average cell brightness resulting in the increase of the descriptor length from 3780 to 3908 values, but it is easy to compute and instantly gives ≈ 25% improvement of the miss rate at 10‒4 False Positives Per Window (FPPW). The modification has been tested on two versions of HOG-based descriptors: the classic Dalal-Triggs and the modified one, where, instead of spatial Gaussian masks for blocks, an additional central cell has been used. The proposed modification is suitable for hardware implementations of HOG-based detectors, enabling an increase of the detection accuracy or resignation from the use of some hardware-unfriendly operations, such as a spatial Gaussian mask. The results of testing its influence on the brightness changes of test images are also presented. The descriptor may be used in sensor networks equipped with hardware acceleration of image processing to detect humans in the images. |
topic |
digital image processing object detection human detection |
url |
http://www.degruyter.com/view/j/mms.2016.23.issue-1/mms-2016-0012/mms-2016-0012.xml?format=INT |
work_keys_str_mv |
AT wojcikowskimarek histogramoforientedgradientswithcellaveragebrightnessforhumandetection |
_version_ |
1724749649468719104 |