Unsupervised Segmenting Focused Objects in Low Depth of Field Images

K.D. Zhang, H.Q. Lu, M.Y. Duan, and Q. Zhao (PRC)


Image segmentation, Low depth of field, Focused object, Wavelet modulus maxima, Mean shift


Unsupervised segmenting region of interest in images is very useful in image indexing for content-based retrieval and target recognition. The proposed method is designed to separate the focused object from blurred foreground or background in low depth of field (DOF) images automatically. The proposed method is based on the shift of mean value in combination with detecting the edges using wavelet modulus maxima. Experiments testify that the focused object regions can be segmented with the method acceptably.

Important Links:

Go Back