Muhittin Gökmen

License Plate Character Segmentation Based on the Gabor Transform and Vector Quantization

This paper presents a novel algorithm for license plate detection and license plate character segmentation problems by using the Gabor transform in detection and local vector quantization in segmentation. As of our knowledge this is the first application of Gabor filters to license plate segmentation problem. Even though much of the research efforts are devoted to the edge or global thresholding-based approaches, it is more practical and efficient to analyze the image in certain directions and scales utilizing the Gabor transform instead of error-prone edge detection or thresholding. Gabor filter response only gives a rough estimate of the plate boundary. Then binary split tree is used for vector quantization in order to extract the exact boundary and segment the plate region into disjoint characters which become ready for the optical character recognition.

Tracking and Segmentation of Highway Vehicles in Cluttered and Crowded Scenes

Monitoring highway traffic is an important application of computer vision research. In this paper, we analyze congested highway situations where it is difficult to track individual vehicles in heavy traffic because vehicles either occlude each other or are connected together by shadow. Moreover, scenes from traffic monitoring videos are usually noisy due to weather conditions and/or video compression. We present a method that can separate occluded vehicles by tracking movements of feature points and assigning over-segmented image fragments to the motion vector that best represents the fragments movement. Experiments were conducted on traffic videos taken from highways in Turkey, and the proposed method can successfully separate vehicles in overpopulated and cluttered scenes.

Robust Face Alignment for Illumination and Pose Invariant Face Recognition

In building a face recognition system for real-life scenarios, one usually faces the problem that is the selection of a feature-space and preprocessing methods such as alignment under varying illumination conditions and poses. In this study, we developed a robust face alignment approach based on Active Appearance Model (AAM) by inserting an illumination normalization module into the standard AAM searching procedure and inserting different poses of the same identity into the training set. The modified AAM search can now handle both illumination and pose variations in the same epoch, hence it provides better convergence in both point-to-point and point-to-curve senses. We also investigate how face recognition performance is affected by the selection of feature space as well as the proposed alignment method. The experimental results show that the combined pose alignment and illumination normalization methods increase the recognition rates considerably for all feature-spaces.

Local Zernike Moment Representation for Facial Affect Recognition.

Local representations became popular for facial affect recognition as they efficiently capture the image discontinuities, which play an important role for interpreting facial actions. We propose to use Local Zernike Moments (ZMs) [4] due to their useful and compact description of the image discontinuities and texture. Their main advantage in comparison to well-established alternatives such as Local Binary Patterns (LBPs) [5], is their flexibility in terms of the size and level of detail of the local description. We introduce a local ZM-based representation which involves a non-linear encoding layer (quantisation). The functionality of this layer is mapping similar facial configurations together and increasing compactness. We demonstrate the use of the local ZM-based representation for posed and naturalistic affect recognition on standard datasets, and show its superiority to alternative approaches for both tasks. Contemporary representations are often designed as frameworks consisting of three layers [2]: (Local) feature extraction, non-linear encoding and pooling. Non-linear encoding aims at enhancing the relevance of local features by increasing their robustness against image noise. Pooling describes small spatial neighbourhoods as single entities, ignoring the precise location of the encoded features, and increasing the tolerance against small geometric inconsistencies. In what follows, we describe the proposed local ZM-based representation scheme in terms of this threelayered framework. Feature Extraction – Local Zernike Moments: The computation of (complex) ZMs can be considered equivalent to representing an image in an alternative space. As shown in Figure 1-a, an image is decomposed onto a set of basis matrices (ZM bases), which are useful for describing the variation at different directions and scales. ZM bases are orthogonal, therefore there is no overlap in the information conveyed by each feature (ZM coefficient). ZMs are usually computed for the entire image, however in this case, ZMs cannot capture the local variation due to ZM bases lacking localisation [3]. In contrary, when computed around local neighbourhoods across the image, they become an efficient tool for describing the image discontinuities which are essential to interpreting facial activity. Non-linear Encoding – Quantisation: We perform quantisation via converting local features into binary values. Such coarse quantisation increases compactness and allows us to code each local block only with a single integer. Figure 1-b illustrates the process of obtaining the Quantised Local ZM (QLZM) image. Firstly, local ZM coefficients are computed across the input image (LZM layer) — each image in the LZM layer (LZM image) contains the features that are extracted through a particular ZM basis. Next, each LZM image is converted into a binary image by quantising each pixel via the signum(·) function. Finally, the QLZM image is obtained by combining all of the binary images. Specifically, each pixel in a particular location of the QLZM image is an integer (QLZM integer), computed by concatenating all of the binary values in the corresponding location of all binary images. The QLZM image is similar to an LBP-transformed image, in the sense that it contains integers of a limited range. Yet, the physical meaning of the information encoded by each integer is quite different. LBP integers describe a circular block by considering only the values along the border, neglecting the pixels that remain inside the block. Therefore, the efficient operation scale of LBPs is usually limited to 3-5 pixels [1, 5]. QLZM integers, on the other hand, describe blocks as a whole, and provide flexibility in terms of operation scale without major loss of information. Pooling – Histograms: Our representation scheme pools encoded features over local histograms. Figure 1-c illustrates the overall pipeline of the proposed representation scheme. Firstly, the QLZM image is computed through the process that is illustrated in detail in Figure 1-b. Next, . . . . . . ... = ZM Coefficients (local features) ZM Bases

An Active Illumination and Appearance (AIA) Model for Face Alignment

Face recognition systems are typically required to work under highly varying illumination conditions. This leads to complex effects imposed on the acquired face image that pertains little to the actual identity. Consequently, illumination normalization is required to reach acceptable recognition rates in face recognition systems. In this paper, we propose an approach that integrates the face identity and illumination models under the widely used active appearance model framework as an extension to the texture model in order to obtain illumination-invariant face localization.

Traffic sign recognition using Scale Invariant Feature Transform and color classification

In this paper, we propose a traffic sign detection and recognition technique by augmenting the scale invariant feature transform (SIFT) with new features related to the color of local regions. SIFT finds local invariant features in a given image and matches these features to the features of images that exist in the training set. Recognition is performed by finding out the training image that gives the maximum number of matches. In this study, performance of SIFT in traffic sign detection and recognition issue is investigated. Afterwards, new features which increase the performance are added. Those are color inspection by using proposed color classification method and inspecting the orientations of SIFT features. These features check the accuracy of matches which are found by SIFT. Color classification method finds out true colors of the pixels by applying some classification rules. It is observed that adding color and orientation inspections raises the recognition performance of SIFT significantly. Obtained results are very good and satisfying even for the images containing traffic signs which are rotated, have undergone affine transformations, have been damaged, occluded, overshadowed, had alteration in color, pictured in different weather conditions and different illumination conditions.

Local Zernike Moments: A new representation for face recognition

In this paper, we propose a new image representation called Local Zernike Moments (LZM) for face recognition. In recent years, local image representations such as Gabor and Local Binary Patterns (LBP) have attracted great interest due to their success in handling difficulties of face recognition. In this study, we aim to develop an alternative representation to further improve the face recognition performance. We achieve this by utilizing Zernike Moments which have been successfully used as shape descriptors for character recognition. We modify global Zernike moments to obtain a local representation by computing the moments at every pixel of a face image by considering its local neighborhood, thus decomposing the image into a set of images, moment components, to capture the micro structure around each pixel. Our experiments on FERET face database reveal the superior performance of LZM over Gabor and LBP representations.

Super-resolution reconstruction of faces by enhanced global models of shape and texture

We present a computationally efficient method for the super-resolution reconstruction of face images from their low-resolution versions. It is based on generative models and utilizes both the shape and texture components together. The main idea is that the image details can be synthesized by global modeling of accurately aligned local image regions. In order to achieve sufficient accuracy in alignment, shape reconstruction is considered as a separate problem and solved together with texture reconstruction in a coordinated manner. Meanwhile, the statistical dependency between the shape and texture components is also considered. Moreover, different from traditional model-based super-resolution methods, we use a corrected form of the degradation operator with the aligned images. We show that when the degradation is used with the aligned texture components as is, it causes bias in the reconstructions. To overcome this problem, we reflect the same processing performed in alignment onto the degradation operator and use this corrected version in texture reconstruction. Experimental results show that the proposed solution provides superior image reconstructions (both qualitatively and quantitatively) in a faster way.

Concurrent Segmentation and Recognition with Shape-Driven Fast Marching Methods

We present a variational framework that integrates the statistical boundary shape models into a Level Set system that is capable of both segmenting and recognizing objects. Since we aim to recognize objects, we trace the active contour and stop it near real object boundaries while inspecting the shape of the contour instead of enforcing the contour to get a priori shape. We get the location of character boundaries and character labels at the system output. We developed a promising local front stopping scheme based on both image and shape information for fast marching systems. A new object boundary shape signature model, based on directional Gauss gradient filter responses, is also proposed. The character recognition system that employs the new boundary shape descriptor outperforms the other systems, based on well-known boundary signatures such as centroid distance, curvature etc

Image Compression Based on Centipede Model

We present an efficient contour based image coding scheme based on Centipede Model. Unlike previous contour based models which presents discontinuities with various scales as a step edge of constant scale, the centipede model allows us to utilize the actual scales of discontinuities as well as location and contrast across them. The use of the actual scale of edges together with other properties enables us to reconstruct a better replica of the original image as compared to the algorithm lacking this feature. In this model, there is a centipede for each edge segment which lies along the segment and the gray level variation across an edge point is represented by the difference between footholds and distance between left and right feet of the centipede. We obtain edges by using the recently introduced Generalized Edge Detector (GED) [1] which controls the scale and shape of the filter, providing edges suitable to the application in hand. The detected edge segments are ranked based on the weighted sum of the length of the segment, mean contrast and standard deviation of gray values on the segment. In our scheme, the compression ratio is controlled by retaining the most significant segments and by adjusting the distance between the successive foot pairs. The original image is reconstructed from this sparse information by minimizing a hybrid energy functional which spans a space called Λτ-space. Since the GED filters are derived from this energy functional, we utilized the same process for detecting the edges and reconstructing the surface from them. The proposed model and the algorithm have been tested on both real and synthetic images. Compression ratio reaches to 180:1 for synthetic images while it ranges from 25:1 to 100:1 for real images. We have experimentally shown that the proposed model preserves perceptually important features even at the high compression ratios.