Shape Feature Analyse And Retrieval Of Image

Content-based image retrieval (CBIR) is one of the most active hotspots in the fields of image engineering. It can directly analyze image content such as color, shape and texture, etc., describe image features reasonably and make retrieval process more effective and more adaptive to human vision. However the shape is an important character which can distinguish an object from the other objects,and it is one of the most prominent character of image. shape-based image retrieval is one of most important part of content-based image retrieval,by analyzing the shape feature,we can find the more effective representation of the images.The main contents and contributions of this thesis are summarized as follows:(1) The key techniques of shape-based image retrieval, which concludes shapefeature analyse and retrieval and similarity measurement in feature space, are studied in depth. The presented techniques are summarized; also the advantages and disadvantages of them are analyzed from the view of application.(2) A method which based on the axis of least inertia is modified for image matching and retrieval.It is modified by the axis of least inertia and component relationship among the feature points. It is invariant with respect to translation,rotation,reflection and scaling, and robust to minor deformations and occlusions.This method have good performance and retrieval efficiency.(3)Compared the polygon matching and retrieval methods which have presented,present a similarity measurement which by combining the axis of least inertia and chain and decompose the polygon to the triangles.By comparing this method with other several methods,the result indicates it is more efficient.