Leaf area measurement based on Markov random field

This paper concerns the plant leaf area measurement based on improved image processing. Firstly, the referenced rectangle was detected with 2-side scan method. Then the leaf region was segmented according to 2G-R-B of every pixel with two thresholds, and by using of dilatation operation, the trimap of leaf image was got. Next the pixels in unknown area were classified to the foreground or background area with improved knockout method and the exact leaf was segmented. Lastly, the leaf area was calculated according to the pixels proportion between leaf region and the referenced rectangle. Experiment results show our methods have good accuracy and rapid speed.

[1]  David Salesin,et al.  A Bayesian approach to digital matting , 2001, Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition. CVPR 2001.

[2]  Qian Tian,et al.  Matlab Image Processing Technique and Application in Pore Structure Characterization of Hardened Cement Pastes , 2013 .

[3]  Chaur-Chin Chen,et al.  Color images' segmentation using scale space filter and markov random field , 1992, Pattern Recognit..

[4]  Dani Lischinski,et al.  A Closed-Form Solution to Natural Image Matting , 2006, IEEE Transactions on Pattern Analysis and Machine Intelligence.