WaSS: A Novel Hybrid Method for Object Recognition Using Wavelet based Statistical and Structural Approaches
Object recognition is one of the most important tasks in computer vision domain. In this paper, a novel method is proposed to recognize objects using shape information based on statistical and structural approaches. For the extraction of shape information, first decompose the original image, then real complement approach using rotations are proposed. Wavelets rearrange the shape of an object for reaching a desirable size. In addition, a set of statistical features are constructed, which can be used for object recognition. We have applied this method to representation and recognition of Flavia and Swedish leaf datasets. Experiments show that a combined statistical and structural approach is superior to other state-of-the-art methods.
Keywords: Object recognition, shape, statistical features, real complement approach.
Download Full-Text
ABOUT THE AUTHORS
G.Karuna
Research Scholar GRIET, Hyderabad
B.Sujatha
professor, Dept. of CSE GIET, Rajahmundry
P.Chandrasekhar Reddy
Professor Dept. of ECE JNTUH, Hyderabad
G.Karuna
Research Scholar GRIET, Hyderabad
B.Sujatha
professor, Dept. of CSE GIET, Rajahmundry
P.Chandrasekhar Reddy
Professor Dept. of ECE JNTUH, Hyderabad