Golnaz Abdollahian

A Low Complexity Sign Detection and Text Localization Method for Mobile Applications

We propose a low complexity method for sign detection and text localization in natural images. This method is designed for mobile applications (e.g., unmanned or handheld devices) in which computational and energy resources are limited. No prior assumption is made regarding the text size, font, language, or character set. However, the text is assumed to be located on a homogeneous background using a contrasting color. We have deployed our method on a Nokia N800 cellular phone as part of a system for automatic detection and translation of outdoor signs. This handheld device is equipped with a 0.3-megapixel camera capable of acquiring images of outdoor signs that typically contain enough details for the sign to be readable by a human viewer. Our experiments show that the text of these images can be accurately localized within the device in a fraction of a second.

Neighborhood matching for object recognition algorithms based on local image features

Local image features around interest-points have been widely used in order to exploit the similarities between different views of an object in different images. While there are numerous algorithms on detecting the interest-points and defining the local features, few have focused on the importance of the matching process. In this paper, we presented a method that matches interest-points detected via any algorithm. The method is motivated from human perceptual rules, particularly the Gestalt Psychology, and realizes the fact that “The whole is different from the sum of its parts”. The efficacy of the algorithm is not only the ability to decrease the number of false positive matches but also to increase the number of true positives, yielding rock-steady results for any algorithm based on matching local features.

Finding Regions of Interest in Home Videos Based on Camera Motion

In this paper, we propose an algorithm for identifying regions of interest (ROIs) in video, particularly for the keyframes extracted from a home video. The camera motion is introduced as a new factor that can influence the visual saliency. The global motion parameters are used to generate location-based importance maps. These maps can be combined with other saliency maps calculated using other visual and high-level features. Here, we employed the contrast-based saliency as an important low level factor along with face detection as a high level feature in our approach.

A study on the effect of camera motion on human visual attention

The aim of this paper is to examine the effect of camera motion in user generated video with respect to human visual attention. Having a more accurate human attention model is particularly useful in applications such as video summarization where the identification of visual importance is crucial to the quality of the results. Most models proposed thus far, have not considered camera motion as an independent factor in identifying visual saliency. In this study eye movement was recorded while subjects watched videos with different types of camera motion. The distribution of the fixation points indicated high correlation between the visual saliency and the type and direction of camera movement. The statistical tests confirmed that the results are statistically significant.

Spatial and temporal models for texture-based video coding

In this paper, we investigate spatial and temporal models for texture analysis and synthesis. The goal is to use these models to increase the coding efficiency for video sequences containing textures. The models are used to segment texture regions in a frame at the encoder and synthesize the textures at the decoder. These methods can be incorporated into a conventional video coder (e.g. H.264) where the regions to be modeled by the textures are not coded in a usual manner but texture model parameters are sent to the decoder as side information. We showed that this approach can reduce the data rate by as much as 15%.

User generated video annotation using Geo-tagged image databases

In this paper we propose a system that annotates a user generated video based on the associated location metadata, by exploiting user-tagged image databases. An example of such a database is a photo sharing website such as Flickr [1] where users upload their images and annotate them with various tags. The goal is to find the tags that have high probability of being relevant to the video without any complex object or action recognition being done to the video sequence. A video is first segmented into camera views and a set of keyframes are selected to represent the video. We will describe the concept of camera view as the basic element of user generated videos which has special properties suitable for the video annotation application. The keyframes are used to retrieve the most relevant images in the database. A “tag processing” step is then used to tag the video.

A non-parametric adaptive algorithm for the censored newsvendor problem

This article studies the problem of determining stocking quantities in a periodic-review inventory model when the demand distribution is unknown. Moreover, lost sales are unobservable in the system and hence inventory decisions are to be made solely based on sales data. Both the non-perishable and perishable inventory problems are addressed. Using an online convex optimization procedure, a non-parametric adaptive algorithm that produces inventory policy in each period that depends on the entire history of stocking decisions and sales observations. With the help of a convex quadratic underestimator of the cost function, it is established that the T-period average expected cost of the inventory policy converges to the optimal newsvendor cost at the rate of O(log T/T) for demands whose expected cost functions satisfy an α-exp-concavity property. It is shown that, when the demand distribution is continuous, this property holds the probability density function over the decision set is bounded away from zero. For other continuous distributions, a “shifted” version of the density function is constructed to show an ε-consistency property of the algorithm so that the gap between the T-period average expected cost of the proposed policy and the optimal newsvendor cost is of the order O(log T/T) + ε (for a given small ε > 0). Simulation results show that the proposed algorithm performs consistently better than two existing algorithms that are closely related to the proposed algorithms.

A curvicylindrical coordinate system for the visualization and segmentation of the ascidian tail

State of the art biological imaging methods, such as confocal microscopy, create 3D volumes by sampling on a cartesian grid. This cartesian coordinate system is often not convenient for visualization and analysis of multi layered organs or tissues. The ascidian embryonic tail, for example, is organized along anterioposterior (AP), dorsoventral (DV) and left-right (LR) axes that are locally orthogonal but curved in the XYZ microscope space. Here, we propose a “curvicylindrical” coordinate system for analysis of such biological structures. By extracting representative paths that traverse different tissue layers, the embryo can be visualized in a small number of 2D images (3 images in the case of the ascidian tail). As we demonstrate, this reduction of the dimensionality from 3D to 2D facilitates the initialization process for high quality segmentation of different cell types, and identification of tissue boundaries.

A region-dependent image matching method for image and video annotation

In this paper we propose an image matching approach that selects the method of matching for each region in the image based on the region properties. This method can be used to find images similar to a query image from a database, which is useful for automatic image and video annotation. In this approach, each image is first divided into large homogeneous areas, identified as “texture areas”, and non-texture areas. Local descriptors are then used to match the keypoints in the non-texture areas, while texture regions are matched based on low level visual features. Experimental results prove that while exclusion of texture areas from local descriptor matching increases the efficiency of the whole process, utilization of appropriate measures for different regions can also increase the overall performance.

A low complexity method for detection of text area in natural images

We propose a low complexity method for segmentation of text regions in natural images. This algorithm is designed for mobile applications (e.g. unmanned or hand-held devices) in which computational and energy resources are limited. No prior assumption is made regarding the text size, font, language, character set or the camera angle. However, the text is assumed to be located on a piecewise homogeneous background with a contrasting color. We have deployed our method on a Nokia N800 Internet tablet as part of a system for automatic detection and translation of outdoor signs. Our experiments show that the 0.3 megapixel images taken by the phone camera can be accurately segmented within the device in a fraction of a second.