Background Subtraction for Dynamic Texture Scenes Using Local FCH features With Adaptive Updating Procedure

A conventional color histogram (CCH) considers neither the color similarity across different bins nor the color dissimilarity in the same bin. Therefore, it is sensitive to noisy interference such as illumination changes and quantization errors. Furthermore, CCHs large dimension or histogram bins requires large computation on histogram comparison. To address these concerns, this paper presents a new color histogram representation, called fuzzy color histogram (FCH), by considering the color similarity of each pixel's color associated to all the histogram bins through fuzzy-set membership function. A novel and fast approach for computing the membership values based on fuzzy c-means algorithm is introduced. The proposed FCH is further exploited in the application of image indexing and retrieval. Experimental results clearly show that FCH yields better retrieval results than CCH. Such computing methodology is fairly desirable for image retrieval over large image databases.

Content-Based Image Retrieval (CBIR) Systems are becoming increasingly important and finding application in diverse areas of discipline. Due to disturbing factors present in images, e.g. non-uniform illumination and lightening intensities, the retrieval results are not always satisfactory. In this paper, we discuss how fuzzy set theory can be used to formulate an efficient image signature. This signature is 1-D Fuzzy Color Histogram (FCH) consisting of a small number of bins. The signature is obtained from image contents by considering the contribution of each pixels color to all the histogram bins through the use of fuzzy-sets membership functions. Experimental results on a database of over 10,000 images demonstrate that the proposed system is less sensitive light intensity changes and more robust than the Conventional Color Histogram (CCH) in retrieval precision.

Dynamic backgrounds and sudden illumination changes are two of the major problems associated with background subtraction techniques. In this paper, we present a novel approach to background subtraction that addresses both of these challenges. In particular, we present an improved codebook background modelling and subtraction technique. We utilise image segmentation on the background image and model the background with a codebook for each pixel along with a pseudo background layer. We perceive background motion as an occlusion of one background layer by a nearby background layer. In other words, sliding of one background layer over a neighbouring layer causes background motion and will hence result in false segmentation. We present our approach of codeword spreading across layer boundaries to handle background motion. Furthermore, we present a two-step update of the background codebook to handle both sudden and gradual illumination changes.

For the image classification task, the color histogram is widely used as an important color feature indicating the content of the image. However, the high-resolution color histograms are usually of high dimension and contain much redundant information which does not relate to the image content, while the low-resolution histograms cannot provide adequate discriminative information for image classification. In this paper, a new color feature representation is proposed which not only takes the correlation among neighbouring components of the conventional color histogram into account but removes the redundant information as well. A high-resolution, uniform quantized color histogram is first obtained from the image. Then the redundant bins are removed and some neighbouring bins are combined together to generate a new feature component to maximize the discriminative ability. The mutual information is adopted to evaluate the discriminative power of a specific feature set and an iterative algorithm is performed to derive the histogram quantization and their corresponding feature generation. To illustrate the effectiveness of the proposed feature representation, an application of detecting adult images, i.e., image classification between erotic and benign images, is carried out. Two widely used classification techniques, SVM and Adaboost, are employed as the classifier. Experimental results show the superior performance of our color representation compared with the conventional color histogram in image classification.

Effect of multi filters in glucoma detection using random forest classifier

A new algorithm is proposed for background subtraction in highly dynamic scenes. Background subtraction is equated to the dual problem of saliency detection: background points are those considered not salient by suitable comparison of object and background appearance and dynamics. Drawing inspiration from biological vision, saliency is defined locally, using center-surround computations that measure local feature contrast. A discriminant formulation is adopted, where the saliency of a location is the discriminant power of a set of features with respect to the binary classification problem which opposes center to surround. To account for both motion and appearance, and achieve robustness to highly dynamic backgrounds, these features are spatiotemporal patches, which are modeled as dynamic textures. The resulting background subtraction algorithm is fully unsupervised, requires no training stage to learn background parameters, and depends only on the relative disparity of motion between the center and surround regions. This makes it insensitive to camera motion. The algorithm is tested on challenging video sequences, and shown to outperform various state-of-the-art techniques for background subtraction.

Background subtraction is the algorithmic process that segments out the region of interest often known as foreground from the background. Extensive literature and numerous algorithms exist in this domain, but most research have focused on videos captured by static cameras. The proliferation of portable platforms equipped with cameras has resulted in a large amount of video data being generated from moving cameras. This motivates the need for foundational algorithms for foreground/background segmentation in videos from moving cameras. In this dissertation, I propose three new types of background subtraction algorithms for moving cameras based on appearance, motion, and a combination of them. Comprehensive evaluation of the proposed approaches on publicly available test sequences show superiority of our system over state-of-the-art algorithms. The first method is an appearance-based global modeling of foreground and background. Features are extracted by sliding a fixed size window over the entire image without any spatial constraint to accommodate arbitrary camera movements. Supervised learning method is then used to build foreground and background models. This method is suitable for limited scene scenarios such as Pan-Tilt-Zoom surveillance cameras. The second method relies on motion. It comprises of an innovative background motion approximation mechanism followed by spatial regulation through a Mega-Pixel denoising process. This work does not need to maintain any costly appearance models and is therefore appropriate for resource constraint ego-vision systems. The proposed segmentation combined with skin cues is validated by a novel application on authenticating hand-gestured signature captured by wearable cameras. The third method combines both motion and appearance. Foreground probabilities are jointly estimated by motion and appearance. After the mega-pixel denoising process, the probability estimates and gradient image are combined by Graph-Cut to produce the segmentation mask. This method is universal as it can handle all types of moving cameras.

Using color histograms in automatic emotion recognition systems faces different issues. One of the important challenges is to determine the appropriate number of bins in the color histogram to achieve the highest recognition performance possible with minimal computations. This research focuses on emotion recognition induced by visual contents of images, or REVC for short, using ARTphoto dataset. Twenty-two different classifiers are used with color histograms in both RGB (red, green, blue) and HSV (hue, saturation, value) color spaces across different numbers of bins, and overall performance of each bin size is compared with that of other bin sizes. The research findings show that the performance of REVC system does not improve in terms of overall sensitivity rate, when the number of bins in color histogram is increased. Moreover, this paper identifies the advantage of using HSV color space over RGB in using color histogram for REVC systems. Furthermore, findings recognize the optimum number of bins in both RGB and HSV color spaces, and ANOVA (analysis of variance) is used to analyze experimental data, which identifies the optimum color histogram bin size used for HSV color space is significantly better than that used for RGB color space in REVC systems.

Sensor devices are regularly used on unmanned aerial vehicles (UAVs) as reconnaissance and intelligence gathering systems and as support for front line troops on operations. This platform provides a wealth of sensor data and has limited computational power available for processing. The objective of this work is to detect and identify objects in real-time, with a low power footprint so that it can operate on a UAV. An appraisal of current computer vision methods is presented, with reference to their performance and applicability to the objectives. Experimentation with real-time methods of background subtraction and motion estimation was carried out and limitations of each method described. A new, assumption free, data driven method for object detection and identification was developed. The core ideas of the development were based on models that propose that the human vision system analyses edges of objects to detect and separate them and perceives motion separately, a function which has been modelled here by optical flow. The initial development in the temporal domain combined object and motion detection in the analysis process. This approach was found to have limitations. The second iteration used a detection component in the spatial domain that extracts texture patches based on edge contours, their profile, and internal texture structure. Motion perception was performed separately on the texture patches using optical flow. The motion and spatial location of texture patches was used to define physical objects. A clustering method is used on the rich feature set extracted by the detection method to characterise the objects. The results show that the method carries out detection and identification of both moving and static objects, in real-time, irrespective of camera motion.

We introduce a new approach for salient motion in dynamic scenes. The proposed method is based on dynamic texture (DT) model and observability of the linear systems. Each video clip of τ frames is represented by DT in a holistic manner and the learned DT parameters are used to form an efficient formula of measuring observability. The formula is related to time-domain eigenvalues and eigenvectors. But the eigenvalue decomposition operation is not needed. The salient motion is detected by thresholding the observability values of each pixel location. Besides, we also introduce the system property of observability which can be used to reduce computational cost. Our method is tested on a challenging sequences set. Experimental results show that the proposed method has better trade-off between detection results and time efficiency than most current methods.

Object detection in video surveillance is typically done through background subtraction or temporal differencing. While these techniques perform very well under scenes where there are minimal light changes, they begin to fail when the scene contains rapid illumination changes. The effects of this is most profound in indoor environments. Under these conditions, the background modeling techniques produce large numbers of false positives. This paper proposes a sequential approach to suppressing these false positives. Both frame and region level spatial scales are considered to detect sudden light changes and make use of Gabor filter responses and edge maps to identify and remove false positives.

The timely identification of dermatological diseases together with their precise diagnosis enables proper medical interventions that minimize further disease complications. The authors conduct a performance- based evaluation of AdaBoost classifier with handcrafted image features for automated skin disease classification systems. The researchers worked with a specific collection of 38,000 dermatological pictures which included ten different disease types from Melanoma to Eczema to Psoriasis to Fungal Infections. The research investigated six image feature extraction methods including Gabor Filter and JPEG Coefficient Filter (JPEGCF) and Pyramid Histogram of Oriented Gradients Filter (PHOGF) as well as Simple Color Histogram Filter (SCHF) and Fuzzy Color and Texture Histogram Filter (FCTH) with Fuzzy Opponent Histogram Filter (FOHF). The research used feature vectors extracted from AdaBoost classifiers in WEKA within 10-fold cross-validation procedures for training purposes. The evaluation metrics consisted of accuracy, precision, recall, ROC, PRC as well as training time. Within the examined models Gabor+AdaBoost demonstrated maximum rates of accuracy (97.78%) alongside precision (0.98) and recall (0.98) but it required the most execution time at 14.4s. When weighed against each other JPEGCF+AdaBoost maintained equivalent accuracy at 97.59% but required only 0.21 seconds to complete tasks thus earning status as a balanced choice for this system. SCHF+AdaBoost proved suitable for real-time operations by delivering 94.11% accuracy results within 0.06 seconds computing time. The study reveals that Gabor descriptors provide optimal predictive accuracy but JPEGCF along with SCHF demonstrate the best combination of performance with computational speed.

Falling snow has the visual property that it is simultaneously a set of discrete moving particles as well as a dynamic texture. To capture the dynamic texture properties of falling snow using particle systems can, however, require so many particles that it severely impacts rendering rates. Here we address this limitation by rendering the texture properties directly. We use a standard particle system to generate a relatively sparse set of falling snow flakes, and we then composite in a dynamic texture to fill in between the particles. The texture is generated using a novel image-based spectral synthesis method. The spectrum of the falling snow texture is defined by a dispersion relation in the image plane, derived from linear perspective. The dispersion relation relates image speed, image size, and particle depth. In the frequency domain, it relates the wavelength and speed of moving 2D image sinusoids. The parameters of this spectral snow can be varied both across the image and over time. This provides the flexibility to match the direction and speed parameters of the spectral snow to those of the falling particles. Camera motion can also be matched. Our method produces visually pleasing results at interactive rendering rates. We demonstrate our approach by adding snow effects to static and dynamic scenes. An extension for creating rain effects is also presented.

SUMMARYThis paper proposes a person search system that uses clothing features. The system uses a camera to obtain the appearance of a person and a range sensor for person detection and direction estimation. The system detects a moving person based on background subtraction using distance information and histogram of oriented gradient (HOG) features. The clothing area is extracted according to the positions of the person's shoulders and waist. The clothes are characterized by four features that are calculated from the color histogram, color differences, frequency, and co‐occurrence matrix. The system recognizes the same clothing based on the degree of similarity, which is obtained by comparing clothing features. However, two problems can be assumed: (1) the appearance of clothing differs depending on the body direction and (2) the color of clothing changes depending on the illumination. For the former problem, we prepared three databases (front/side/rear views) corresponding to the direction of the person. The system estimates the direction of the person by using information on the body arc based on ellipse fitting. The direction is determined as perpendicular to the long axis of the ellipse fitted to the cross‐section of the body. Concerning the latter problem, we calibrated the color information to correspond to the environment. The method recognizes clothes by selecting the database provided for the direction of the person. The results showed that it was robust in identifying specific persons.

Real-time foreground object extraction is an important subject for computer vision applications. Model-based background subtraction methods have been used to extract the foreground objects. Different from previous methods, this paper introduces a hybrid codebook-based background subtraction method by combining the mixture of Gaussian (MOG) with the codebook (CB) method. We propose an ellipsoid CB model for modeling the dynamic background with highlight and shadow, and develop a modified shadow/ highlight removal method to overcome the influence of illumination change. Our method can avoid extracting the false foreground pixels (e.g., dark background) or missing the real foreground pixels (e.g., bright foreground). Finally, we have done two experiments to compare the performance of our method with the others based on [18] and the change detection benchmark dataset provided in CVPR 2011, respectively.