Object detection is a key step in several computer vision applications. Some of these include 3-D reconstruction, compression, medical imaging, augmented reality, image retrieval, and surveillance applications. It is especially important for visual surveillance sensor nodes as it segments the input image into background and foreground. All following steps analyze only foreground parts of the image; background is discarded. This chapter reviews the main approaches used for surveillance purposes and focuses on background subtraction techniques. It also presents a Hybrid Selective scheme based on Mixture of Gaussian modeling, named HS-MoG. The scheme provides accurate results for outdoor scenes with clutter motion in the background while requiring fewer operations than the original Mixture of Gaussian scheme. © 2014 Springer Science+Business Media, LLC.
CITATION STYLE
Al Najjar, M., Ghantous, M., & Bayoumi, M. (2014). Object detection. Lecture Notes in Electrical Engineering, 114, 97–118. https://doi.org/10.1007/978-1-4614-1857-3_5
Mendeley helps you to discover research relevant for your work.