The proposed metric is to use latent quality aware topics in an image to measure blurriness. A novel image quality vocabulary is firstly obtained by the contrast features computed from the training images using K-means. Probabilistic latent semantic analysis model is then used to discover quality aware topics that are latent in clear sample images and the test image. The similarity between the latent topics of the test image and the average topics of clear images is finally computed to measure blurriness. Experimental results show that the proposed blur metric is monotonic, robust to additive noises, and also consistent with the human visual system.

The segmentation and recognition modules are usually implemented sequentially in most traditional automatic license recognition (LPR) systems. In this work, we integrate segmentation and recognition into a Markov network, where bidirectional constraints between segmentation and recognition are exploited for LPR. In addition, both low-level structural attributes and compositional semantics of license plates are incorporated in a probabilistic way. A belief propagation (BP) algorithm is used for statistical inference that is able to separate and recognize license characters simultaneously. Experiments on Chinese license plates show that the proposed approach work well even when connected and distorted characters present.

In this paper, we address the issue of extracting contour of the object with a specific shape. A hierarchical graphical model is proposed to represent shape variations. A complex shape is decomposed into several components which are described as principal component analysis (PCA) based models in various levels. The hierarchical representation allows for chain-like conditional dependency within a single level and bidirectional communication between different levels. Additionally, a sequential Monte-Carlo (SMC) based inference algorithm that can explore the graphical structure is proposed to estimate the contour. The experiments performed on real-world hand and face images show that the proposed method is effective in combating occlusion and cluttered background. Moreover, it is possible to isolate the localization error to an individual component of a shape attributed to the hierarchical representation.

In practice, we meet numerous image restoration problems, particularly those on face images. An approach to image restoration based on statistical prior of faces and image blur measure is presented in this paper. The novelty of the approach is twofold: 1) the prior on the shape and appearance of a face represented as statistical models is incorporated into the regularized image restoration formulation. Moreover, an iterative algorithm is given to provide a numerical solution. 2) An image blur measure is defined to describe the blur nature of an image. The measure is evaluated at each step of the iteration process to determine the weight of the constraint introduced by the statistical prior. Both subjective and objective comparisons of the restoration results indicate great improvements on image preservation and noise suppression.