Yanwen Chong

A new pedestrian detection method based on combined HOG and LSS features

Abstract Pedestrian detection is a critical issue in computer vision, with several feature descriptors can be adopted. Since the ability of various kinds of feature descriptor is different in pedestrian detection and there is no basis in feature selection, we analyze the commonly used features in theory and compare them in experiments. It is desired to find a new feature with the strongest description ability from their pair-wise combinations. In experiments, INRIA database and Daimler database are adopted as the training and testing set. By theoretic analysis, we find the HOG–LSS combined feature have more comprehensive description ability. At first, Adaboost is regarded as classifier and the experimental results show that the description ability of the new combination features is improved on the basis of the single feature and HOG–LSS combined feature has the strongest description ability. For further verifying this conclusion, SVM classifier is used in the experiment. The detection performance is evaluated by miss rate, the false positives per window, and the false positives per image. The results of these indicators further prove that description ability of HOG–LSS feature is better than other combination of these features.

Gene selection using independent variable group analysis for tumor classification

Microarrays are capable of detecting the expression levels of thousands of genes simultaneously. So, gene expression data from DNA microarray are characterized by many measured variables (genes) on only a few samples. One important application of gene expression data is to classify the samples. In statistical terms, the very large number of predictors or variables compared to small number of samples makes most of classical “class prediction” methods unemployable. Generally, this problem can be avoided by selecting only the relevant features or extracting new features containing the maximal information about the class label from the original data. In this paper, a new method for gene selection based on independent variable group analysis is proposed. In this method, we first used t-statistics method to select a part of genes from the original data. Then, we selected the key genes from the selected genes for tumor classification using IVGA. Finally, we used SVM to classify tumors based on the key genes selected using IVGA. To validate the efficiency, the proposed method is applied to classify three different DNA microarray data sets. The prediction results show that our method is efficient and feasible.

Integrated real-time vision-based preceding vehicle detection in urban roads

This paper presents a real-time algorithm for a vision-based preceding vehicle detection system. The algorithm contains two main components: vehicle detection with various vehicle features, and vehicle detection verification with dynamic tracking. Vehicle detection is achieved using vehicle shadow features to define a region of interest (ROI). After utilizing methods such as histogram equalization, ROI entropy and mean of edge image, the exact vehicle rear box is determined. In the vehicle tracking process, the predicted box is verified and updated. Test results demonstrate that the new system possesses good detection accuracy and can be implemented in real-time operation.

MutualCascade method for pedestrian detection

Abstract An effective and efficient feature selection method based on Gentle Adaboost (GAB) cascade and the Four Direction Feature (FDF), namely, MutualCascade, which can be applied to the pedestrian detection problem in a single image, is proposed in this paper. MutualCascade improves the classic method of cascade to remove irrelevant and redundant features. The mutual correlation coefficient is utilized as a criterion to determine whether a feature should be chosen or not. Experimental results show that the MutualCascade method is more efficient and effective than Voila and Jones’ cascade and some other Adaboost-based method, and is comparable with HOG-based methods. It also demonstrates a higher performance compared with the state-of-the-art methods.

Eigenface-based sparse representation for face recognition

Face recognition has been a challenging task in computer vision. In this paper, we propose a new method for face recognition. Firstly, we extract HOG (Histogram of Orientated Gradient) features of each class face images in used Face databases. Then, we select the so-called eigenfaces from HOG features corresponding to each class face images and finally use them to build a overcomplete dictionary for ESRC (the Eigenface-based Sparse Representation Classification ). Experiments show that our method receives better results by comparison.

Distributed Storage Algorithm for Geospatial Image Data Based on Data Access Patterns

Declustering techniques are widely used in distributed environments to reduce query response time through parallel I/O by splitting large files into several small blocks and then distributing those blocks among multiple storage nodes. Unfortunately, however, many small geospatial image data files cannot be further split for distributed storage. In this paper, we propose a complete theoretical system for the distributed storage of small geospatial image data files based on mining the access patterns of geospatial image data using their historical access log information. First, an algorithm is developed to construct an access correlation matrix based on the analysis of the log information, which reveals the patterns of access to the geospatial image data. Then, a practical heuristic algorithm is developed to determine a reasonable solution based on the access correlation matrix. Finally, a number of comparative experiments are presented, demonstrating that our algorithm displays a higher total parallel access probability than those of other algorithms by approximately 10–15% and that the performance can be further improved by more than 20% by simultaneously applying a copy storage strategy. These experiments show that the algorithm can be applied in distributed environments to help realize parallel I/O and thereby improve system performance.

A Global User-Driven Model for Tile Prefetching in Web Geographical Information Systems

A web geographical information system is a typical service-intensive application. Tile prefetching and cache replacement can improve cache hit ratios by proactively fetching tiles from storage and replacing the appropriate tiles from the high-speed cache buffer without waiting for a client’s requests, which reduces disk latency and improves system access performance. Most popular prefetching strategies consider only the relative tile popularities to predict which tile should be prefetched or consider only a single individual users access behavior to determine which neighbor tiles need to be prefetched. Some studies show that comprehensively considering all users’ access behaviors and all tiles’ relationships in the prediction process can achieve more significant improvements. Thus, this work proposes a new global user-driven model for tile prefetching and cache replacement. First, based on all users’ access behaviors, a type of expression method for tile correlation is designed and implemented. Then, a conditional prefetching probability can be computed based on the proposed correlation expression mode. Thus, some tiles to be prefetched can be found by computing and comparing the conditional prefetching probability from the uncached tiles set and, similarly, some replacement tiles can be found in the cache buffer according to multi-step prefetching. Finally, some experiments are provided comparing the proposed model with other global user-driven models, other single user-driven models, and other client-side prefetching strategies. The results show that the proposed model can achieve a prefetching hit rate in approximately 10.6% ~ 110.5% higher than the compared methods.

Low-rank and eigenface based sparse representation for face recognition.

In this paper, based on low-rank representation and eigenface extraction, we present an improvement to the well known Sparse Representation based Classification (SRC). Firstly, the low-rank images of the face images of each individual in training subset are extracted by the Robust Principal Component Analysis (Robust PCA) to alleviate the influence of noises (e.g., illumination difference and occlusions). Secondly, Singular Value Decomposition (SVD) is applied to extract the eigenfaces from these low-rank and approximate images. Finally, we utilize these eigenfaces to construct a compact and discriminative dictionary for sparse representation. We evaluate our method on five popular databases. Experimental results demonstrate the effectiveness and robustness of our method.

Pedestrian detection based on gradient and texture feature integration

In this paper, based on feature integration, we proposed a new method for pedestrian detection. Firstly, we extracted the histogram of oriented gradients (HOG) feature and local binary pattern (LBP) feature from the original images respectively. Secendly, K-singular value decomposition (K-SVD) was used to extract sparse representation features from the HOG and LBP features. Moreover, PCA was used to reduce the dimension of HOG and LBP. Finally, we combined the PCA based features and the K-SVD based sparse representation features directly for fast pedestrian detection in still images. Experimental results on two databases show that the proposed approach is effective for pedestrian detection.

Feature Selection and Pedestrian Detection Based on Sparse Representation

Pedestrian detection have been currently devoted to the extraction of effective pedestrian features, which has become one of the obstacles in pedestrian detection application according to the variety of pedestrian features and their large dimension. Based on the theoretical analysis of six frequently-used features, SIFT, SURF, Haar, HOG, LBP and LSS, and their comparison with experimental results, this paper screens out the sparse feature subsets via sparse representation to investigate whether the sparse subsets have the same description abilities and the most stable features. When any two of the six features are fused, the fusion feature is sparsely represented to obtain its important components. Sparse subsets of the fusion features can be rapidly generated by avoiding calculation of the corresponding index of dimension numbers of these feature descriptors; thus, the calculation speed of the feature dimension reduction is improved and the pedestrian detection time is reduced. Experimental results show that sparse feature subsets are capable of keeping the important components of these six feature descriptors. The sparse features of HOG and LSS possess the same description ability and consume less time compared with their full features. The ratios of the sparse feature subsets of HOG and LSS to their full sets are the highest among the six, and thus these two features can be used to best describe the characteristics of the pedestrian and the sparse feature subsets of the combination of HOG-LSS show better distinguishing ability and parsimony.