The design and implementation of pedestrian detection system for stationary camera
碩士 === 國立高雄應用科技大學 === 電子工程系碩士班 === 103 === Recently, human detection has been an important issue and has been widely used in many applications. Histograms of oriented gradients (HOG) is proven to offer better feature extraction that could significantly outperform existing feature sets for human dete...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | zh-TW |
| Published: |
2015
|
| Online Access: | http://ndltd.ncl.edu.tw/handle/vsx47g |
| Summary: | 碩士 === 國立高雄應用科技大學 === 電子工程系碩士班 === 103 === Recently, human detection has been an important issue and has been widely used in many applications. Histograms of oriented gradients (HOG) is proven to offer better feature extraction that could significantly outperform existing feature sets for human detection. When we adopt HOG with SVM to do human detection in the images or videos, sliding window and progressive scan mode is used. It takes a lots of windows and the detecting time is long.
In this thesis, we present the preprocessing skills for fixed camera in HOG-based human detection system. Compared with the traditional HOG-based human detection system, our system achieves faster speed and better detection.
|
|---|