多目標物件辨識之研究

Multi-object Detection

彭憲彰(Xian Zhang Peng)

指導教授 ： 李錫捷

元智大學/資訊學院/資訊管理學系/碩士(2008年)

摘要

本論文描述使用機器學習的方法在物件偵測的領域中，使在影像處理的過程中能非常的快速且高命中率，首先利用Haar-like特徵對偵測器做快速的運算，接著使用命為Gentle adaboost的學習演算法，在由Haar-Like所產生的特徵下，選取少許但關鍵的視覺化特徵，並由此特徵產生許多微弱但有效的分類器。

接著把許多微弱的產生器結合在一起，使成為一個強大的分類器cascade，主要是讓影像中的背景被快速的偵測出來且被丟棄，然後在對於可能是所要偵測之物件做更多的運算已確認是否為該物件。

關鍵字

哈爾相似特徵、機器學習、分類器

並列摘要

This study describes a machine learning approach for visual object detection which is capable of processing images rapidly and achieving high detection rates。

The proposed detector computes the Haar-Like Features at the first stage. In the second stage, a learning algorithm called Gentle Adaboost is used which selects a small number of critical visual features and yields extremely efficient classifiers.

The final stage of the object detection process is to combine those complex classifiers into something called “Cascade” which allows background regions of the image to be quickly discarded while spending most computation efforts on regions with candidate objects.

並列關鍵字

Haar-like feature、Machine Learning、Classifier

參考文獻

[2] H. Schneiderman, T. Kanade, “A statistical method for 3D object detection applied to faces and cars”, in: IEEE Conference on Computer Vision and Pattern Recognition, IEEE, 2000.

[3] N.H. Bergboer *, E.O. Postma, H.J. van den Herik,”Context-based object detection in still images”, Image and Vision Computing, Volume 24, Issue 9, vol.1,September 2006, Pages 987-1000.

[4] C. Papageorgiou, M. Oren, and T. Poggio. “A general framework for Object Detection”,In International Conference on Computer Vision, 1998.

detection in images by components”,IEEE Transactions on Pattern

[6] Paul Viola and Michael J. Jones,“Rapid Object Detection using a Boosted Cascade of Simple Features”,IEEE CVPR, 2001.