Showing papers on "Illumination problem published in 2009"
TL;DR: In this article, an adaptive ellipse approach (AEA) was proposed to identify the mild-rust-color spots properly using an Adaptive Ellipse (AE) to overcome the non-uniform illumination problem for image segmentation.
Abstract: Image processing has been used for assessment of infrastructure surface coating conditions for years. In North America, civil engineers have utilized image recognition for steel bridge coating inspection since late 1990s. However, there is still no robust method to overcome the non-uniform illumination problem for infrastructure surface coating defect recognition to date. Therefore, this paper aims to develop a new approach to tackle the non-uniform illumination problem for rust image recognition. This paper starts with an investigation of 14 color spaces in order to find out the best color configuration for non-uniformly illuminated rust image segmentation. Then, the identified best color configuration a⁎b⁎, which has a moderate ability to filter light, is utilized to develop the proposed adaptive ellipse approach (AEA). In AEA, a rust image is partitioned into three parts: background, rust, and mild-rust-color spots. The main idea is to identify the mild-rust-color spots properly using an adaptive ellipse. Illumination adjustment is also adopted in this approach to overcome the non-uniform illumination problem. Finally, the performance of the AEA-based a⁎b⁎ configuration is compared to the K-Means method, one of the most popular and effective image recognition approaches, to show the effectiveness of the proposed AEA approach.
••01 Nov 2009
TL;DR: The existing methods for illumination problem in face recognition are classified into three main categories and the representative algorithms and theories are introduced, then the advantages and disadvantages of these algorithms correspondingly are analyzed.
Abstract: The change of illumination problem is still a challenging and difficult problem in face recognition under complex illumination condition. There were many proposals that dealing with illumination problem in face recognition in the past decade. In this paper, we classify the existing methods into three main categories and introduce the representative algorithms and theories, then analyze the advantages and disadvantages of these algorithms correspondingly. As a result of this study we conclude direction of future research.
04 Jun 2009
TL;DR: Two Empirical Mode Decomposition (EMD) based face recognition schemes are proposed in this paper to address variant illumination problem and the experimental results on the PIE database verify the efficiency of the proposed methods.
Abstract: Two Empirical Mode Decomposition (EMD) based face recognition schemes are proposed in this paper to address variant illumination problem. EMD is a data-driven analysis method for nonlinear and non-stationary signals. It decomposes signals into a set of Intrinsic Mode Functions (IMFs) that containing multiscale features. The features are representative and especially efficient in capturing high-frequency information. The advantages of EMD accord well with the requirements of face recognition under variant illuminations. Earlier studies show that only the low-frequency component is sensitive to illumination changes, it indicates that the corresponding high-frequency components are more robust to the illumination changes. Therefore, two face recognition schemes based on the IMFs are generated. One is using the high-frequency IMFs directly for classification. The other one is based on the synthesized face images fused by high-frequency IMFs. The experimental results on the PIE database verify the efficiency of the proposed methods.
TL;DR: The results show that in terms of the fluctuations contrast amplitudes in an image, the local-global method give better results than the standard homomorphic filtering technique.
Abstract: Variable illumination in a texture is a common problem occurs to a real-time image modalities. The imbalance illumination in a texture creates virtual regions within one image, hence it affects the performance of the classification methods because it introduced an artifact patterns or virtual regions to an image. This paper presents a method to overcome the variable illumination problem in a pig skin texture using the information in the local and global blocks. The focus of this paper is to provide a fast, reliable and safe method to stabilize the lighting in an image. Pig skin texture is selected because it has a special pattern characteristic that needs to be preserved. The results show that in terms of the fluctuations contrast amplitudes in an image, the local-global method give better results than the standard homomorphic filtering technique.
24 Nov 2009
TL;DR: A novel method for solving nonuniform illumination problem using multiresolution decomposition and a new technique called hillcreast-valley classification with adaptive mean filter to normalize illumination and detect dominant facial features, such as eyes, nose and mouth automatically.
Abstract: Automatic facial feature detection is one of the most important topics in computer vision and there are still many open problems that have not been solved. Nonuniform illumination is among one of those problems. This paper proposes a novel method for solving nonuniform illumination problem using multiresolution decomposition and a new technique called hillcreast-valley classification with adaptive mean filter to normalize illumination and detect dominant facial features, such as eyes, nose and mouth automatically. The proposed method is divided into three modules: eye detection, nose detection, and mouth detection modules. In this method, a single face image is divided into three regions: eye, nose, and mouth regions, then we use multiresolution decomposition to detect the eyes, and use thresholding to detect the nose and the mouth. For multiresolution decomposition, we decompose the eye region into small blocks and use hillcrest-valley classification with adaptive mean filter to classify each block as either a high or low-intensity region. Each low-intensity(valley) region is then decomposed into smaller blocks and each block is classified as either high- or low-intersity region. The low-intensity regions are then defined as the eyes. Finally the nose and the mouth are detected using thresholding. The method was evaluated on the YaleB face database that consists of face images taken by different illumination variations and the experimental results indicate that our proposed method achieves high accuracy rate.
TL;DR: A nonlinear manifold framework for the face pose and the face illumination normalization processing is proposed and the efficient face tracking and recognition results on indoor and outdoor video are derived.
Abstract: Face tracking and recognition are difficult problems because the face is a non-rigid object. The main reasons for the failure to track and recognize the faces are the changes of a face pose and environmental illumination. To solve these problems, we propose a nonlinear manifold framework for the face pose and the face illumination normalization processing. Specifically, to track and recognize a face on the video that has various pose variations, we approximate a face pose density to single Gaussian density by PCA(Principle Component Analysis) using images sampled from training video sequences and then construct the GMM(Gaussian Mixture Model) for each person. To solve the illumination problem for the face tracking and recognition, we decompose the face images into the reflectance and the illuminance using the SSR(Single Scale Retinex) model. To obtain the normalized reflectance, the reflectance is rescaled by histogram equalization on the defined range. We newly approximate the illuminance by the trained manifold since the illuminance has almost variations by illumination. By combining these two features into our manifold framework, we derived the efficient face tracking and recognition results on indoor and outdoor video. To improve the video based tracking results, we update the weights of each face pose density at each frame by the tracking result at the previous frame using EM algorithm. Our experimental results show that our method is more efficient than other methods.Keywords:Face recognition, Face tracking, Manifold, Retinex