Local structures of shadow boundaries as well as complex interactions of image regions remain largely unexploited by previous shadow detection approaches. In this paper, we present a novel learning-based framework for shadow region recovery from a single image. We exploit local structures of shadow edges by using a structured CNN learning framework. We show that using structured label information in classification can improve local consistency over pixel labels and avoid spurious labelling. We further propose and formulate shadow/bright measure to model complex interactions among image regions. The shadow and bright measures of each patch are computed from the shadow edges detected by the proposed CNN. Using the global interaction constraints on patches, we formulate a least-square optimization problem for shadow recovery that can be solved efficiently. Our shadow recovery method achieves state-of-the-art results on major shadow benchmark databases collected under various conditions.

In many intelligent surveillance systems there is a requirement to search for people of interest through archived semantic labels. Other than searching through typical appearance attributes such as clothing color and body height, information such as whether a person carries a bag or not is valuable to provide more relevant targeted search. We propose two novel and fast algorithms for sling bag and backpack detection based on the geometrical properties of bags. The advantage of the proposed algorithms is that it does not require shape information from human silhouettes therefore it can work under crowded condition. In addition, the absence of background subtraction makes the algorithms suitable for mobile platforms such as robots. The system was tested with a low resolution surveillance video dataset. Experimental results demonstrate that our method is promising.

Background subtraction is an essential technique for automatic video analysis. The main idea is to construct and update the model of the background scene. Foreground pixels are detected if they deviate from the background model to a certain extent. The model can consist of color, texture and gradient information [1]. In this paper, we focus on both color and texture information. The proposed texture feature is based on local binary pattern (LBP), while the color feature is represented by local color pattern (LCP). LBP is known to work well on texture rich regions and is invariant to subtle illumination variations, but it is inefficient on uniform regions. In view of this, color information can be incorporated to complement the texture feature. On the other hand, when the scene contrast or video quality is poor, color information may be unreliable and should be assigned lower priority than texture information. We propose a fuzzy rule-based system that adaptively adjusts the weights of the texture and color features based on the pixel's local properties. Experimental results on real scenes demonstrate the robustness of the proposed method.

Texture and color are two primitive forms of features that can be used to describe a scene. While conventional local binary pattern (LBP) texture based background subtraction performs well on texture rich regions, it fails to detect uniform foreground objects in large uniform background. As such, color information can be used to complement texture feature. In this study, we propose to incorporate local color feature based on Improved Hue, Luminance, and Saturation (IHLS) color space and introduce an adaptive scheme that automatically adjusts the weight between texture and color similarities based on the pixel's local properties: texture uniformity and color saturation. Experiments on eight challenging sequences demonstrate the effectiveness of the proposed method compared to the state-of-the-art algorithms.

This paper presents a new keypoint-based approach for Near-Duplicate (ND) keyframes identification using global and local features. In the proposed method, the matched keypoints are filtered for affine transform estimation based on an affine invariant feature which is using the lesser matching pairs and voting bins. To further confirm the matching, the Local Binary Pattern (LBP) and color features of areas formed by matched keypoints in two images are compared. This method combines the advantage of global and local matching, it is able to identify two images related with affine transform. The proposed algorithm has been tested in Columbia dataset and conducted the quantitative comparison with RANSAC algorithm and SR-PE algorithm. The experiment results show that the proposed method can achieve good results.

Shape and motion are two most distinct cues observed from human actions. Traditionally, K-Nearest Neighbor (K-NN) classifier is used to compute crisp votes from multiple cues separately. The votes are then combined using linear weighting scheme. Usually, the weights are determined in a brute-force or trial-and-error manner. In this study, we propose a new classification framework based on sum-rule fusion of fuzzy K NN classifiers. Fuzzy K-NN classifier is capable of producing soft votes, also known as fuzzy membership values. Based on Bayes theorem, we show that the fuzzy membership values produced by the classifiers can be combined using sum-rule. In our experiment, the proposed framework consistently outperforms the conventional counterpart (K-NN with majority voting) for both Weizmann and KTH datasets. The improvement may attribute to the ability of the proposed framework to handle data ambiguity due to similar poses present in different action classes. We also show that the performance of our method compares favorably with the state-of-the-arts.

Most methods for multiple camera tracking rely on accurate calibration to associate data from multiple cameras. However, it often is not easy to have an accurate calibration in some real applications due to practical reasons. The inaccurate calibration can lead to wrong data association of objects between cameras. In this paper, we propose a method to handle the data association of objects in multiple cameras under inaccurate ground plane homography by using the RFS Bayes filter. Our method is based on modeling measurements from cameras to a random finite set. This random finite set includes the primary measurement from the object, extraneous measurements of the object, and clutter. Experimental results show the efficiency and robustness of our method through challenging cases such as occlusions, merged and split persons.

The performances of conventional crisp and fuzzy K-nearest neighbor (K-NN) algorithms trained using finite samples tends to be poor . With ldquoholesrdquo in the training data, it is unlikely that the decision area formed can actually represent the underlying data distribution. There is a need to capture more useful information from the limited training samples, therefore we propose a new fuzzy rule-based K-NN algorithm. A fuzzy rule-based initialization procedure differentiates our proposed algorithm from the conventional fuzzy K-NN algorithm. The new initialization procedure allows us to handle the imprecise inputs (neighborhood density and distance) through the natural framework of fuzzy logic system. Unlike conventional K-NN algorithms, the ability to fine tune the membership functions can lead to a highly versatile decision boundary. Thus, the new algorithm can be specifically tuned for different problems to achieve better results. The advantage is demonstrated on a synthetic data set in two-dimensional space. In addition, we also adopt weighted Euclidean distance measurement to overcome the curse of dimensionality . The Euclidean distance weights and the parameters of the fuzzy rule-based system are then optimized with genetic algorithm (GA) simultaneously. The practical applicability of the proposed algorithm is verified on four UCI data sets (Bupa liver disorders, Glass, Pima Indians diabetes and Wisconsin breast cancer) and Ford automotive data set with an improvement of 3.42% in classification rate on average.

This paper presents a hybrid TSK fuzzy rule-based classifier. Fuzzy c-means clustering and Genetic algorithm and are used to optimize the number of rules and antecedent parameters. By using the relationship between a SVM and a TSK FLS, an efficient method for learning the consequent parts of the TSK fuzzy system is introduced. The resulting hybrid fuzzy classifier has a compact rule base and good generalization capabilities compared to existing algorithms in the literature. In this sense, the curse of dimensionality which is often associated with fuzzy rule-based classifier can be avoided. The performance of the proposed hybrid fuzzy classifier is verified through extensive tests and comparison with other methods.

This paper studies the ability of a non-singleton fuzzy logic system (NSFLS) that is evolved using Genetic Algorithm (GA) to handle the uncertainties in pattern classification problems. The performance of non-singleton and singleton systems for cardiac arrhythmias classification is compared. Results show that NSFLS can deal with uncertainty within its framework more efficiently, thereby enabling classification to be performed using features that are easier to extract.

This paper aims at assessing the feasibility of using a type-2 fuzzy system for ECG arrhythmic beat classification. Three types of ECG signals, namely the normal sinus rhythm (NSR), ventricular fibrillation (VF) and ventricular tachycardia (VT), are considered. The inputs to the fuzzy classifier are the average period and the pulse width, two features that are commonly used for computer-assisted arrhythmia recognition and are readily extracted from pre-processed ECG waveforms. Using a combination of the fuzzy C-means clustering algorithm and the amount of dispersion in each cluster, a method for designing the antecedent type-2 MFs of the classifier from a training data set is formulated. Tests using data from the MIT-BIH Arrhythmia Database show that the proposed type-2 fuzzy classifier yields an accuracy of 90.91 % for VT events and 84 % for VF events and 100 % for NSR events.