A Novel 3D Model Retrieval Method Based on the Object Outline of Image

To compensate the drawbacks of traditional 3D model retrieval methods which adopt several keywords or a sample model, the paper proposes a novel retrieval method based on the image which contains user's desired object. Since object's outline can efficiently describe its content especially its shape, we first extract the object outline from image and then obtain the set of shape points from the outline. Through comparing the shape feature of the outlines of 3D model's multiple 2D projections with that of the desired object's outline, the related 3D models are returned. The experiments conducted on a classical 3D model database, Princeton Shape Benchmark, show that our method can efficiently retrieve the desired models according to the images obtained from web and digital camera.