Maneuvering Target Tracking in Cluttered Background Based on Color Invariance and Support Vector Machine

We present a reversible data hiding technique for digital medical images. The idea is to embed the header of the image or other data in the image itself, pursuing tamper-proofing capabilities and inseparability. In particular, the proposed data hiding technique using Least Significant Bits (LSBs) hides information into the Region of Interest (ROI) and preserves lost information in the Region of Non Interest (RONI) for reversibility purposes. ROI pixels are accessed consecutively or in accordance to a Just Noticeable Distortion (JND) model, while RONI pixels are accessed in a consecutive or pseudo-random order. In the experiments with MRI and X-Ray images, we prove excellent fidelity and 100% reversibility of the ROI.

Scene classification from images is a challenging problem in computer vision due to its significant variability of scale, illumination, and view. Recently, Latent Dirichlet Allocation (LDA) model has grown popular in computer vision field, especially in scene labeling and classification. However, the effectiveness of the LDA model for the scene classification has not yet been addressed thoroughly. Especially, there is little experimental evaluation on the model's performance for different types of features. Fusion of multiple types of features is usually necessary in the scene classification due to the complexity of scene images. In this paper, we investigate the effectiveness of the LDA model in scene classification by using 7 types of features (i.e. uniform grid based interest points, Harris corner based interest points, scale invariant feature transform (SIFT), texture, shape, color, and location) and their various combinations. Furthermore, we compare the performance of the LDA model with Support Vector Machine (SVM) classifier. All experiments are performed on the UIUC Sport Scene database. The experiments demonstrate that the performance of the LDA model 1) is significantly lower than the SVM classifier for the scene classification over different types of features; and 2) decreases by fusing multiple features while improvement shown in SVM classifier.

This paper proposes a robust watermarking method against high‐density salt and pepper noise attacks. Automatic region of interest (ROI) detection, embedding encoded data, removing different densities of noise, data extraction by omitting, and labeling the noisy pixels of Region of Non‐Interest (RONI), and decoding the extracted data using ROI pixel information as a key, are various steps of the presented scheme. The automatic ROI detection method separates the RONI from ROI with four vertices of the smallest rectangle, for the embedding process. The encoded watermark data is embedded into the least significant bits of RONI in four neighbour pixels. Adaptive Removal of high‐density Salt and pepper Noise method can enhance image quality and reduce the effect of salt and pepper noise attacks. The embedded information is preserved from destruction if the host image is impaired through the power of robustness. The best results are obtained through the action of extracting the watermark from RONI pixels, utilizing the same ROI detection method. Omitting and labeling the noisy pixels of the RONI will ensure healthy extracted watermark data, leading to decreased Bit Error Rate (BER) values. Finally, these data are interpreted using the key of ROI pixels, and the watermark data is decoded and retrieved. Due to salt and pepper noise obliterating pixel bits and their corresponding transform coefficients in the transform domain, the spatial domain is employed to enhance robustness against such attacks. The results show the high performance of the presented scheme. The average BER value for five MRI databases in a 97% salt and pepper noise attack is 38.6.

A simple and effective object recognition scheme is to represent and march images on the basis of color histograms. To obtain robustness against varying imaging circumstances (e.g. a change in illumination, object pose, and viewpoint), color histograms are constructed from color invariants. However in general, color invariants are negatively affected by sensor noise due to the instabilities of these color invariant transforms at many RGB values. To suppress the effect of noise blow-up for unstable color invariant values, in this paper color invariant histograms are computed using variable kernel density estimators. To apply variable kernel density estimation in a principled way, models are proposed for the propagation of sensor noise through color invariants. As a result the associated uncertainty is known for each color invariant value. The associated uncertainty is used to derive the parameterization of the variable kernel density estimator during histogram construction. It is empirically verified that the proposed method compares favorably to traditional color histograms for object recognition.

Segmentation and classification of objects in images is one of the most important and yet one of the most complex problems in computer vision. Many different things contribute to the complexity of this problem, but amongst those that are most significant are variations in object appearance due to various perspectives, changes in scene illumination and typically a large number of partially occluded objects in a scene. Humans have the ability to successfully deal with these complications and to recognize objects in scenes without seemingly much effort. However, to aid their recognition process, humans use different sources of information such as previous experiences, context of objects placed in a scene and rules learned in the past that suggest how the physical world should appear and behave. Computer vision is trying to replicate the same strategies used by human vision and apply them to computer systems with the aim of improving typical tasks of detection, localization and classification of image objects. In this work we propose a new model for natural image object classification using contextual information at the level of image segments. Context modeling is largely independent of appearance-based classification and proposed model enables simple upgrade of existing systems with information from global and/or local context. Context modeling is based on non-parametric use of appearance-based classification results which is a novel approach compared to previous systems that model context on a limited number of rules expressed with a fixed set of parameters. Model implementation resulted in a system that, in our simulations, showed stable improvement of the appearance-based object classification. DOI: http://dx.doi.org/10.5755/j01.itc.46.1.13610

Gender recognition from facial images has become one of challenging research problem in computer vision, security, verbal-nonverbal communication and human computer interaction applications nowadays. Because facial images include many information such as gender, facial expressions, age, ethnic origin in computer-aided applications, the success rate of the gender recognition depends on quality of facial images. In this paper, it is proposed a new gender recognition method combining Speed Up Robust Features (SURF) based Bags of Visual Words (BoW) and Support Vector Machine (SVM) algorithm unlike previous work. The method is tested on realistic frontal, left and right face images from modern gender recognition FERET dataset with 3560 samples to see efficiency of the proposed method. Experimental results show that the proposed method can obtain better gender recognition performance on FERET database and the accuracy level of on left and right face images is a bit lower than the average accuracy level of frontal ones.

Constrained Active Region Models for Fast Tracking in Color Image Sequences

One of the fundamental problems in computer vision is the background-foreground segmentation and most of the strategies have severe drawbacks when working with natural images, where there are extreme conditions such as illumination changes combined with sudden background differences or other noise; moreover if the system has to face real-time restrictions. In this case authors focus on a variation of the U-net architecture to obtain the segmentation of the objects (fishes) in every single frame. The U-net has some interesting properties to explore in the case of image segmentation, such as multi-scale parameter combination. The reported preliminary results, working in a context of a multi-target fish tracking are promising, and envisions an approach that could provide a real-time response to long-lasting experiments using HQ video for multi-target tracking in real-time Computer vision systems.

The task of matching image patches is a fundamental problem in computer vision. When sufficiently textured patches are normalized up to similarity transformation, a simple Normalized Cross Correlation (NCC) of corresponding patches will give a high value. In practice, using it on patches per se may not perform well due to the noisy variations of pixel intensities. A more prudent approach will be to apply it to the abstract features extracted by a deep convolutional network. We study the applicability of an NCC based convolutional network for the task of Patch Matching. Further, there may be cases where the network may fail due to insufficient textures. In those cases, a simple pixel difference based method will be beneficial. To this end, we propose to improve the two basic architectures, Siamese networks and Central-Surround stream networks, using robust matching layers for learning the similarities of patches, assisted by a simple cross-entropy loss function. We empirically verify the performance of the proposed models on the challenging UBC Patches dataset and show that they are close to the state-of-the-art. Further, we evaluate their resilience to large illumination changes in two experimental scenarios: 1) by manually varying the patches of UBC Patches by an affine model 2) by using the publicly available Webcam dataset. We demonstrate that our models are indeed very resilient to illumination variations; they reduce the false positive rate to nearly 10%, and improve over the popular methods by nearly 5%. Further, we demonstrate the generalisability of the proposed NCC based matching layer by applying it to Face Recognition and show that it improves the performances of well known networks on a real-world, surveillance dataset.

Finding color representations that are stable to illuminant changes is still an open problem in computer vision. Until now, most approaches have been based on physical constraints or statistical assumptions derived from the scene, whereas very little attention has been paid to the effects that selected illuminants have on the final color image representation. The novelty of this paper is to propose perceptual constraints that are computed on the corrected images. We define the category hypothesis, which weights the set of feasible illuminants according to their ability to map the corrected image onto specific colors. Here, we choose these colors as the universal color categories related to basic linguistic terms, which have been psychophysically measured. These color categories encode natural color statistics, and their relevance across different cultures is indicated by the fact that they have received a common color name. From this category hypothesis, we propose a fast implementation that allows the sampling of a large set of illuminants. Experiments prove that our method rivals current state-of-art performance without the need for training algorithmic parameters. Additionally, the method can be used as a framework to insert top-down information from other sources, thus opening further research directions in solving for color constancy.

Novel view synthesis is a classic problem in computer vision. It refers to the generation of previously unseen views of a scene from a set of sparse input images taken from different viewpoints. One example of novel view synthesis is the interpolation of views in between the two images of a stereo camera. Another classic problem in computer vision is video frame interpolation, which is important for video processing. It refers to the generation of video frames in between existing ones and is commonly used to increase the frame rate of a video or to match the frame rate to the refresh rate of the monitor that the video is being displayed on. Interestingly, off-the-shelf video frame interpolation can directly be employed to successfully perform view interpolation to address the aforementioned stereo view interpolation problem. Video frame interpolation can be seen as temporal novel view synthesis. However, this perspective is usually not considered and novel view synthesis generally concerns generating unseen views in space rather than time. For this reason, the set of sparse input images that is used for spatial novel view synthesis is commonly either captured at the same time, or it is assumed that the scene is static. This paradigm limits the applicability of novel view synthesis in real-world scenarios though. This thesis addresses three applications of novel view synthesis and provides practical solutions that do not require difficult-to-acquire multi-view imagery: video frame interpolation which performs temporal video-to-video synthesis, synthesizing the 3D Ken Burns effect from a single image which performs spatial image-to-video synthesis, synthesizing video action shots which performs spatiotemporal video-to-video and video-to-image synthesis. These applications not only explore different dimensions of time and space, they also perform novel view synthesis on everyday image and video footage. This is in stark contrast to the large body of existing work which focuses on spatial novel view synthesis while requiring multiple input views that were either captured at the same time or under the assumption of a static scene.

Graph and hypergraph matching are fundamental problems in computer vision, which have been receiving a steadily growing attention in the last two decades due to the recent modern algorithms and models that are both fast and accurate. Graph and hypergraph matching are successful in tackling many real-world tasks that require 2D and 3D feature matching and alignment. Matching is a general problem in vision and graphs have always been quintessential for modeling information that has both local and longer range connections. Therefore, graph matching will continue to influence computer vision and robotics research, crossing over many different mathematical and computational models, including Bayesian networks, conditional and Markov random fields, spectral graph models, and, more recently, deep neural networks. While graph matching is limited to using pairwise relationships, hypergraph matching permits the use of relationships between sets of features of any order. Consequently, it carries the promise to make matching more robust to changes in scale, deformations, and outliers. In this chapter, we present several key ideas, computational models, and optimization methods that introduced in the literature the possibility to learn graph and hypergraph matching in a semi-supervised or even completely unsupervised fashion. We start by describing two methods for solving graph and hypergraph matching based on spectral matching and integer projected fixed point methods, introduced in Chap. 1. Then, based on the same intuition (also discussed in Chap. 1) that correct assignments form a strong cluster of agreements, we follow up by introducing the main approach to learning as it applies to graph and hypergraph matching, as well as the more general Maximum A Posteriori (MAP) inference problem in probabilistic graphical models (MRFs and CRFs). In the last part of the chapter, we demonstrate experimentally the effectiveness of the algorithms presented and, whenever possible, provide extensive comparisons to other top published methods. Moreover, we show that our learning approach is general and can significantly boost the performance of different state-of-the-art matching algorithms in the literature.

This paper presents a novel method for Face Recognition (FR) by applying Support Vector Machine (SVM) in addressing this Computer Vision (CV) problem. The SVM is a capable learning classifier capable of training polynomials, neural networks and RBFs. Singular Value Decomposition (SVD) are used for feature extraction and while SVM for classification. The proposed algorithm is tested on four databases, viz., FERET, FRGC Ver. 2.0, CMU-PIE and Indian Face Database. The singular values are filtered, sampled and classified using Gaussian RBF kernal for SVM. The results are compared with other known methods to establish the advantage of SV-M/D. The recall rate for the proposed system is about 90%. The colossal augmentation is due to the simple but efficient feature extraction and the ability to learn in a high dimensional space.

Template matching is a fundamental problem in computer vision, with many applications. Existing methods use sliding window computation for choosing an image-window that best matches the template. For classic algorithms based on SSD, SAD and normalized cross-correlation, efficient algorithms have been developed allowing them to run in real-time. Current state of the art algorithms are based on nearest neighbor (NN) matching of small patches within the template to patches in the image. These algorithms yield state-of-the-art results since they can deal better with changes in appearance, viewpoint, illumination, non-rigid transformations, and occlusion. However, NN-based algorithms are relatively slow not only due to NN computation for each image patch, but also since their sliding window computation is inefficient. We therefore propose in this paper an efficient NN-based algorithm. Its accuracy is similar (in some cases slightly better) than the existing algorithms and its running time is 43–200 times faster depending on the sizes of the images and templates used. The main contribution of our method is an algorithm for incrementally computing the score of each image window based on the score computed for the previous window. This is in contrast to computing the score for each image window independently, as in previous NN-based methods. The complexity of our method is therefore O(|I|) instead of O(|I||T|), where I and T are the image and the template respectively.

Toward designing intelligent PDEs for computer vision: An optimal control approach