Object tracking via adaptive prediction of initial search point on mobile devices

Description

Common feature tracking algorithms, such as SIFT and SURF, are fairly slow in runtime due to the processing of a large amount of external data. If the object is sufficiently small, outside noise may throw off the object detection device without prior knowledge. Machine learning, especially Markov chains, can use prior knowledge to turn a computationally expensive task into a faster, stochastic one.

This project attempts to implement the mapping portion of SLAM, i.e. to develop a mobile application that could create a planar grid given a known reference, then track an object along that grid. In order to constrain the object eld with a xed camera. the algorithm produced adequately tracks a cube with image re nement. Opportunities exist to create a better template for homography estimation than the first image, which will reduce noise in detecting keypoints, to have a generalized algorithm so a template can be chosen on Android without further tuning, and to further integration of the masking region.

Blind Image Quality Prediction Using Joint Statistics of Gradient Magnitude and Laplacian Features

Network traffic prediction by hybrid model (MLP and ARIMA)

http://www.cs.cmu.edu/~jiyanpan/papers/lncs06.pdf