Projects

Face Detection/Recognition by Gabor Feature and AM-FM Demodulation

Constructed a Gabor filter bank to obtain AM-FM component and then capture the features of each image.
Provided Bayesian classifier for dataset.
Detected human faces from complex background and then recognize each of them.

Eyes and lip detection based on daisy petal Gabor filters and color processing

Developed daisy petal Gabor filters derived from original one to detect more details about image texture. Used watershed algorithm for object segmentation.
Combined with detecting white in eyes and red on the lip to locate the position of eyes and mouth.

HTML5 Icon