Yunhui Yan

Research and Perspective on Local Binary Pattern

In view of the theoretical and practical value of local binary pattern (LBP), the various LBP methods in texture analysis and classification, face analysis and recognition, and other detection applications are reviewed. Firstly, the principle of LBP method is briefly discussed, which mainly analyses the threshold operation, the uniform pattern and rotation invariant pattern in LBP method. Secondly, the texture analysis and classification of the LBP method, face analysis and recognition of the LBP method and other detection applications of the LBP method are particular combed and commented. Finally, the existing important problems of the LBP method are analyzed and the future for the LBP method is pointed out.

Adjacent evaluation of local binary pattern for texture classification

An adjacent evaluation window is constructed to modify threshold scheme of LBP.The adjacent evaluation derives two new descriptors: AECLBP and AELTP.The adjacent evaluation plays an important role in solving sensitivity to noise.The proposed approaches are experimented on five texture databases. This paper presents a novel, simple, yet robust texture descriptor against noise named the adjacent evaluation local binary patterns (AELBP) for texture classification. In the proposed approach, an adjacent evaluation window is constructed to modify the threshold scheme of LBP. The neighbors of the neighborhood center gc are set as the evaluation center ap. Surrounding the evaluation center, we set up an evaluation window and calculate the value of ap, and then extract the local binary codes by comparing the value of ap with the value of the neighborhood center gc. Moreover, this adjacent evaluation method is generalized and can be integrated with the existing LBP variants such as completed local binary pattern (CLBP) and local ternary pattern (LTP) to derive new image features against noise for texture classification. The proposed approaches are compared with the state-of-the-art approaches on Outex and CUReT databases, and evaluated on three challenging databases (i.e. UIUC, UMD and ALOT databases) for texture classification. Experimental results demonstrate that the proposed approaches present a solid power of texture classification under illumination and rotation variations, significant viewpoint changes, and significant large-scale challenging conditions. Furthermore, the proposed approaches are more robust against noise and consistently outperform all the basic approaches in comparison.

Vehicle Classification Using Sparse Coding and Spatial Pyramid Matching

This work adopts sparse coding and spatial pyramid matching to classify the vehicle images. The targets of interest, vehicles in the images, are always degraded in the complex circumstance. Hence, it seems difficult to carry out the classification task by the methods combined gray feature and traditional classifiers. Considering the vehicle image without assignment and complex influence caused by weather, this paper proposes a vehicle classification method based on sparse coding and spatial pyramid matching. First, the proposed method extracts a patch-based sparse feature computed with a discriminate dictionary. With dualizing the sparse feature, the spatial pyramid model is employed to generate a long but sparse feature. At last, SVM with the histogram intersection kernel finishes the ultimate classification task. Diverse from the traditional bag of features model employed to compute the histogram in each level of the spatial pyramid, this paper codes the image patch with a fine learned and discriminate dictionary for a better representation than the gradient-based feature extraction. Fast iteration method on computing the sparse feature ensures the real-time need. Experimental results on the vehicle datasets includes sedan, taxi, van, and truck show the efficiency and accuracy of the proposed method for vehicle classification in practice.

Pair of projections based on sparse consistence with applications to efficient face recognition

Dimension reduction based feature extraction and classification method show significant performance on the high-dimensional face images. The traditional dimension reduction methods learn a projection based on the Fisher criterion or local structure of the face images. This work aims at learning a pair of projection based on sparse consistence which is measured by sparse constraint and label information for efficient face recognition. The first projection maps the original high-dimensional face images into a low-dimensional space where each face is sparse, and the second one which can also be treated as a classifier guides the sparse low-dimensional face images to the right label. The pair of projections is optimized together using alternative update rules efficiently. Due to the discriminant power of sparse face images and the supervised classifier, the proposed algorithm integrates the supervised and unsupervised information and is more efficient than them for face recognition on both learning and classifying. Experimental results on the challenging Extended Yale B, AR, and LFW face image databases demonstrate the proposed algorithm on both accuracy and efficiency. HighlightsAn integrate projection learning based on sparse consistence for FR is proposed.Unsupervised/supervised information are introduced into PPL to extract the discriminant feature.Analysis of complexity and convergence guarantee the efficiency of FR.Experiments on the challenging face datasets validate the accuracy and efficiency of PPL.

Micro Surface Defect Detection Method for Silicon Steel Strip Based on Saliency Convex Active Contour Model

Accurate detection of surface defect is an indispensable section in steel surface inspection system. In order to detect the micro surface defect of silicon steel strip, a new detection method based on saliency convex active contour model is proposed. In the proposed method, visual saliency extraction is employed to suppress the clutter background for the purpose of highlighting the potential objects. The extracted saliency map is then exploited as a feature, which is fused into a convex energy minimization function of local-based active contour. Meanwhile, a numerical minimization algorithm is introduced to separate the micro surface defects from cluttered background. Experimental results demonstrate that the proposed method presents good performance for detecting micro surface defects including spot-defect and steel-pit-defect. Even in the cluttered background, the proposed method detects almost all of the microdefects without any false objects.

Noise robust image matching using adjacent evaluation census transform and wavelet edge joint bilateral filter in stereo vision

The adjacent evaluation census is proposed to improve the robustness of census.Two different and complementary metrics are extracted to reduce ambiguity.The random walk is integrated into aggregation and optimization.WEJBF is employed to eliminate error regions of the original disparity map. Automation application systems based on stereo vision require robust image matching methods to achieve available depth image information. This paper presents a novel noise robust stereo matching using adjacent evaluation census transform and wavelet edge joint bilateral filter. The adjacent evaluation census is firstly proposed to improve the robustness against noise of the census transform. Meanwhile, two different and complementary types of metrics are extracted (the adjacent evaluation census mean and the adjacent evaluation census weighted difference). Moreover, the weighted template is composed of four different directions. Then, to improve the robustness of cost aggregation and disparity optimization, the random walk is integrated into the proposed stereo matching method. Additionally, a disparity map post-processing method named wavelet edge joint bilateral filter is employed to eliminate error regions. An obtained wavelet-based edge image is considered as an important weighted coefficient to guide the post-processing. Experimental results demonstrate that the proposed method presents the best performance of the robustness against noise on the Middlebury dataset. Even in the toughest situation with additive Gaussian noise, our method can still achieve the moderate disparity map. In addition, the wider applicability of the proposed method is demonstrated on the KITTI (i.e., Karlsruhe Institute of Technology (KIT) and Toyota Technological Institute at Chicago (TTI-C)) dataset and some typical real-world sequences.

Intelligent assessment of subsurface cracks in optical glass generated in mechanical grinding process

Abstract Grinding process of optical glass has been reported to be related with the creation of subsurface cracks. However, for the time being, most measurement methods have been depended on human operations. In this paper, an intelligent assessment method based on image processing technique is proposed. Grinding trials proved that, the proposed method can accurately (with the biggest relative error of 3.53%) and quickly (nearly 1.6 seconds per micrographs) recognize and measure the subsurface crack depths. More importantly, the proposed method has good robustness to different-sized images. Besides, the method does not require any input parameters or any adjustment of thresholds, therefore the method does not require any prior knowledge of either mechanical grinding process or brittle material behaviors relating with subsurface cracks. Based on above, the proposed method is expected to be meaningful to both metrology equipment companies and optical glass manufacturers.

Binary coding-based vehicle image classification

Vehicle image classification can describe the visual vehicle with a semantically meaningful category directly. Motivated by its importance, this paper proposes a fast vehicle image classification based on binary coding. As for the vehicle image classification, this paper focuses on the image obtained from the video via analyzing the moving object near the key frames. The proposed method extracts a dense boosting binary feature computed with a boosted binary hash function, and then pools the features in different resolutions. At last, the SVM with spatial pyramid kernel finishes the classification task. In this work, 8 bytes for the feature computed with a hash function that ensures the real-time need. Experimental results on the vehicle datasets includes sedan, taxi, van, and truck show the efficiency and accuracy of the proposed method for vehicle classification in practice.

Detail Enhancement for Infrared Images Based on Propagated Image Filter

For displaying high-dynamic-range images acquired by thermal camera systems, 14-bit raw infrared data should map into 8-bit gray values. This paper presents a new method for detail enhancement of infrared images to display the image with a relatively satisfied contrast and brightness, rich detail information, and no artifacts caused by the image processing. We first adopt a propagated image filter to smooth the input image and separate the image into the base layer and the detail layer. Then, we refine the base layer by using modified histogram projection for compressing. Meanwhile, the adaptive weights derived from the layer decomposition processing are used as the strict gain control for the detail layer. The final display result is obtained by recombining the two modified layers. Experimental results on both cooled and uncooled infrared data verify that the proposed method outperforms the method based on log-power histogram modification and bilateral filter-based detail enhancement in both detail enhancement and visual effect.

Neighborhood Estimated Local Binary Patterns for Texture Classification

A novel texture classification approach based on neighborhood estimated local binary patterns (NELBP) is proposed. In the proposed approach, the local surrounding values of neighborhood estimated are introduced to operate binary patterns. Moreover, two different and complementary descriptors (average-based descriptor and differences-based descriptor) are extracted from local patches. Contrast experiments on Outex database and CUReT database demonstrate that the proposed NELBP is more robust to Gaussian noise than the conventional LBP for texture classification. In addition, the results also show that the combined complementary descriptor playes an important role in texture classification.