The Infona portal uses cookies, i.e. strings of text saved by a browser on the user's device. The portal can access those files and use them to remember the user's data, such as their chosen settings (screen view, interface language, etc.), or their login data. By using the Infona portal the user accepts automatic saving and using this information for portal operation purposes. More information on the subject can be found in the Privacy Policy and Terms of Service. By closing this window the user confirms that they have read the information on cookie usage, and they accept the privacy policy and the way cookies are used by the portal. You can change the cookie settings in your browser.
Image blur and image noise are common distortions during image acquisition. In this paper, we systematically study the effect of image distortions on the deep neural network (DNN) image classifiers. First, we examine the DNN classifier performance under four types of distortions. Second, we propose two approaches to alleviate the effect of image distortion: re-training and fine-tuning with noisy images...
With the dawn of the digital era and online availability of multimedia, cases like copyright infringement, violation of intellectual property rights and breach of privacy are not so rare. Consequently one has to find out the culprit responsible for such illegal actions. This work presents a passive approach based on intrinsic signature of in-camera color filter array (CFA) interpolation to classify...
In most cases image distortions modelled by convolution and additive white noise have unknown model parameters, such as convolution kernel (point spread function — PSF) and noise power. Different methods of blind deconvolution which iteratively approximate PSF use some initial kernel estimation; their performance is sufficiently dependent on the precision of that estimate. Modelling initial PSF as...
Although the mobile head-mounted gaze tracker (HMGT) has gained its great success in human-machine interactions, the real implementation of HMGT still poses several significant challenges. The parallax error and the tedious calibration procedure, as two of these challenges, will be addressed in our proposed two-step calibration method. In the first step, instead of fixating at several pre-defined...
Person re-identification (Re-ID) maintains a global identity for an individual while he moves along a large area covered by multiple cameras. Re-ID enables a multi-camera monitoring of individual activity that is critical for surveillance systems. However, the low-resolution images combined with the different poses, illumination conditions and camera viewpoints make person Re-ID a challenging problem...
This paper presents a fast deblurring algorithm to remove camera motion blur from a single photograph using built-in gyroscopes and strong edge prediction. An inaccurate blur kernel or point spread function (PSF) usually leads to an unsatisfying restored result. Hence, we propose a robust three-phase method for accurate PSF estimation. In the first stage, we utilize the embedded gyroscopes to compute...
The development of automatic nutrition diaries, which would allow to keep track objectively of everything we eat, could enable a whole new world of possibilities for people concerned about their nutrition patterns. With this purpose, in this paper we propose the first method for simultaneous food localization and recognition. Our method is based on two main steps, which consist in, first, produce...
This paper presents a new technique for semi-automatic 2D to 3D stereo video conversion. Our algorithm escapes the scope of traditional depth propagation paradigm based on motion estimation and compensation. First of all, we treat the foreground and background depths separately and then combine them to form a final depth map for each video frame. For the foreground parts, they are first segmented...
This paper presents a scalable multiple GPU architecture for super multi-view (SMV) synthesis using the multi-view video plus depth (MVD) data. SMV synthesis is essential to generate 3D contents for the SMV 3D display with hundred views. SMV 3D display, recently released to support 108 viewpoints, shows the multiplexed result of small viewing interval. Hence, we should synthesize the intermediate...
Stereo matching methods estimate depth information of captured images. One way to estimate accurate depth values is to use the distance information. This method enhances the disparity map by preserving the edge region. In order to preserve the depth discontinuity near the edge region, it uses the distance information as a new weighting value for the matching cost function. However, this method has...
Person re-identification aims to match people across non-overlapping camera views. One of the challenges in re-identification is cross view matching, where the gallery and query data belong to different views. This problem is difficult because the person's appearance varies greatly due to significant viewpoint and poses changes. In this paper, we perform Kernel Canonical Correlation Analysis (KCCA)...
Analysis of near-infrared images has a possibility to simply find vein disease. If super-resolution (SR) techniques improve the quality of near-infrared images with a low signal-to-noise ratio, they could detect abnormal veins at an early stage. Deep convolutional neural networks (DCNNs) as a SR technique were applied to downgraded images, and the effectiveness was investigated. The DCNNs with the...
Parabolic motion cameras are used to obtain better deblurring results of scenes with multiple moving objects. The core of its deblurring process is Iterative Re-weighted Least Squares (IRLS) method. In this paper, we design a hardware accelerator for IRLS flow. The ASIC chip is implemented using TSMC 90 nm technology. It is capable of deblurring a 640 × 480 image captured by a parabolic camera with...
In this paper, we present a fast, non-iterative approach to smooth a noisy input on the Special Euclidean Group, SE(3) manifold. The translational part can be smoothed by a simple Gaussian convolution. We then proposed a novel approach to rotation smoothing. Unlike existing rotation smoothing methods using either iterative optimization methods or stochastic filtering methods, our method allows direct...
Visual localization is the process of finding the location of a camera from the appearance of the images it captures. In this work, we propose an observation model that allows the use of images for particle filter localization. To achieve this, we exploit the capabilities of Gaussian Processes to calculate the likelihood of the observation for any given pose, in contrast to methods which restrict...
Recently, several effective features were proposed for person re-identification, such as Weight Histograms of Overlapping Stripes (WHOS) and Local Maximal Occurrence (LOMO), but it still need to explore new effective feature to improve the precision for person re-identification. So, in this paper, we proposed a new Dual Channel Gradient feature, which can be fused with WHOS and LOMO by directly concatenating...
Accurate calculation of a disparity map in real-time is a challenge in computer vision and autonomous applications due to the expensive computation cost. This paper describes a local temporal stereo method, using normalized cross correlation (NCC) as the matching cost function. To increase the performance in terms of speed without degrading the quality, a down sampling of the disparity map in the...
This paper presents a new lane tracking algorithm for the lane departure warning system without using Kalman filter. The system is capable of extracting the true lane boundaries from all detected lines including noise in the frame and estimates its future position. The new algorithm uses the score mechanism to trace the appearance of lines in previous frames using a score variable which indicates...
An efficient people occupancy detection, tracking, and behavior recognition method is introduced in this study. The problem of monitoring wide field can be achieved using the programmable camera network instated of the typical fixed cameras. In addition, based on the depth image feature, the shape feature of the occupant can be used to the activity recognition more accurately.
Automated Dial Reading (ADR) using image processing is a challenging task that has to deal with the dynamics of real time environment. Literature contains limited research work for ADR that is based on background subtraction, object tracking, and pattern recognition. These methods suffer from dynamic environment such as: varying light intensity, poor resolution, and vibrations in capturing device...
Set the date range to filter the displayed results. You can set a starting date, ending date or both. You can enter the dates manually or choose them from the calendar.