Content-based image retrieval is an important research area in the domain of multimedia information processing, while shape-based image retrieval is one of the main aspects of the content-based image retrieval. Image and video are the most intuitive and visual content in the multimedia. Due to the dramatic increasing of the information in today’s era, the efficient management and rapid indexing of the rapidly expanding multimedia information became an urgent problem. In this thesis, the fundamental theory of the content-based image retrieval and the developing process of the application research are firstly introduced. Then, the content-based image retrieval techniques especially the shape retrieval techniques and its current status are reviewed along with some discussions of the key techniques of image retrieval. Owing to the fact that the shape features are one of the most important features of the image that can describe the object reliably, shape-based image retrieval has been an active research area in image retrieval. The main contributions of this thesis are as follows:Firstly, a new image retrieval algorithm based on the distance-histogram derived from shape contour is proposed. In this algorithm, the one-dimensional Gaussian functions of two different scales are firstly employed respectively for the concave and convex part of the contour to generate a simpler and smoother evolved curve that can capture the main information of the original contour, after which, the skeleton of the contour is extracted with a skeletonization algorithm. Finally, the histogram of the distances between the evolved contour and skeleton is used to describe the shape for the retrieval purpose. The algorithm proposed uses not only the contour that represents the shape of an object from its outer part but also the skeleton that preserves the original object’s topology from its inner part. Experimental results show that this algorithm performs well in robustness to both contour transformations of scaling and rotation and to the noise corruptions of the contour.Secondly, a new image retrieval algorithm based on the Hidden Markov Models (HMM) is also proposed in this thesis. In this algorithm, the object contour is firstly partitioned at points with zero curvature value. Then, four structural features of the symmetry, the orientation, the length, and the linearity are extracted for the partitioned contour segments. For the extracted four structural features, the K-means algorithm is used for clustering purpose so that they can be organized into the observation sequences considered as the input for the HMM. Finally, the HMM is used for the classification and retrieval of the objects. Experimental results show that the proposed algorithm is an efficient method for shape retrieval.