Mohammed Ghazal

UAV-based remote sensing for vegetation cover estimation using NDVI imagery and level sets method

Vegetation is a key component of nearly all global ecosystems, having a role of regulating various biogeochemical cycles in nature, as well as maintaining the energy balance at the earths surface and atmospheric boundary layer. Therefore, it is of wide importance to constantly monitor the changes in vegetation cover and structure, especially those that are due to human influence. Backed up by many global initiatives for vegetation preservation and monitoring, it is necessary to have a system to collect and analyse different vegetation indices and products. One of the most important parameters of calculating many of these indices, is the amount of the vegetation cover in a certain area. This paper proposes a compact system for estimating vegetation in a certain area. The proposed system consists of a small UAV platform with the ability of performing autonomous flights and recording the videos of the ground cover using a camera with a modified infrared filter lens for obtaining the composite NDVI videos. The video is analysed and processed to extract the most significant frames that form a mosaic representing the area recorded by the camera. Finally, a contour is calculated forming the boundary around the area on the mosaic image that contains pixels that represent vegetation or photosynthetic activity. The contour and the area inside the boundary can be used for segmentation and estimation of other vegetation indices.

Mobile-Based Archival and Retrieval of Missing Objects Using Image Matching

It is not uncommon to lose everyday objects outside your home. Currently, there are very few technological resources to help locate lost objects. This paper discusses a multi-platform mobile application that provides a solution to this issue. Lost and found mobile application helps people report lost and found objects through their mobile phones rather than going through the procedure of filling up forms. In addition, it has a backend server that runs an algorithm for object matching using Speeded Up Robust Features (SURF) to match images of lost items with that of found items. The application has three versions: General, Elderly, and Organization. The paper presents the results obtained by implementing the proposed mobile application, as well as the accuracy and performance obtained by testing the image matching algorithm on a set of images representing lost and found objects. Collected results show the robust features of SURF algorithm, as well as its invariance to rotation, which often occurs when taking photographs by mobile phones.

Towards smart wearable real-time airport luggage tracking

In this paper, we propose a smart system for realtime tracking of airport luggage using mobile applications and smartwatches. We track using Kalman-filtered Wi-Fi fingerprints collected by active tags. Information about the flights and association with different luggage pieces is inputted pre-flight using QR codes. Our system uses a smart power management scheme fusing multi-sensor and flight data and assesses the risk of a battery drain to warn the user about the need for recharging. A mobile phone application is used to track the arrival of the luggage allowing passengers to rest after a long flight. We tested our proposed systems impact on the airport wireless network and the random occurrence of different travel delays on the tag energy and found it to be able to efficiently deliver real-time luggage tracking information right to passenger smartwatches.

Deep Learning Role in Early Diagnosis of Prostate Cancer

The objective of this work is to develop a computer-aided diagnostic system for early diagnosis of prostate cancer. The presented system integrates both clinical biomarkers (prostate-specific antigen) and extracted features from diffusion-weighted magnetic resonance imaging collected at multiple b values. The presented system performs 3 major processing steps. First, prostate delineation using a hybrid approach that combines a level-set model with nonnegative matrix factorization. Second, estimation and normalization of diffusion parameters, which are the apparent diffusion coefficients of the delineated prostate volumes at different b values followed by refinement of those apparent diffusion coefficients using a generalized Gaussian Markov random field model. Then, construction of the cumulative distribution functions of the processed apparent diffusion coefficients at multiple b values. In parallel, a K-nearest neighbor classifier is employed to transform the prostate-specific antigen results into diagnostic probabilities. Finally, those prostate-specific antigen–based probabilities are integrated with the initial diagnostic probabilities obtained using stacked nonnegativity constraint sparse autoencoders that employ apparent diffusion coefficient–cumulative distribution functions for better diagnostic accuracy. Experiments conducted on 18 diffusion-weighted magnetic resonance imaging data sets achieved 94.4% diagnosis accuracy (sensitivity = 88.9% and specificity = 100%), which indicate the promising results of the presented computer-aided diagnostic system.

A fast stochastic framework for automatic MR brain images segmentation

This paper introduces a new framework for the segmentation of different brain structures (white matter, gray matter, and cerebrospinal fluid) from 3D MR brain images at different life stages. The proposed segmentation framework is based on a shape prior built using a subset of co-aligned training images that is adapted during the segmentation process based on first- and second-order visual appearance characteristics of MR images. These characteristics are described using voxel-wise image intensities and their spatial interaction features. To more accurately model the empirical grey level distribution of the brain signals, we use a linear combination of discrete Gaussians (LCDG) model having positive and negative components. To accurately account for the large inhomogeneity in infant MRIs, a higher-order Markov-Gibbs Random Field (MGRF) spatial interaction model that integrates third- and fourth- order families with a traditional second-order model is proposed. The proposed approach was tested and evaluated on 102 3D MR brain scans using three metrics: the Dice coefficient, the 95-percentile modified Hausdorff distance, and the absolute brain volume difference. Experimental results show better segmentation of MR brain images compared to current open source segmentation tools.

Modeling of series-connected photovoltaic cells

This paper proposes a new model for series-connected photovoltaic (PV) cells, using a modified one-diode equivalent-circuit model. The PV modules comprise many series-connected cells to generate more electrical power. This modified model starts with the conventional one-diode equivalent-circuit (parallel-connected current source with a diode and a shunt resistance which are connected in series with a series resistance and a load) of PV cells and then proposes a new way of connecting the aforementioned circuit elements. The advantage of the presented modified model, is that it can model the series-connected PV cells by a new representation of one-diode equivalent-circuit. To validate the results of the modified model, similar input variables are applied to the conventional and the presented models. The current/voltage (I/V) characteristics are then calculated from both models and compared. The results show that the difference between the calculated I/V characteristics using the two models is much less than 1 percent. The presented approach can thus, be very useful for researchers or engineers to quickly and easily determine the performance of PV modules.

Smart Mobile-Based Emergency Management and Notification System

Mobile government is an innovative research area where efforts are made to advance governmental and public services. As such, in this paper, we propose a crisis management system for real-time emergency notification of users using mobile applications and smart watches. We develop an intuitive web portal using a client server architecture for governmental agencies to easily and efficiently notify users within the range of danger in the occurrence of a disaster through SMS or push notifications to the mobile application or the smart watch, if the latter is available. Moreover, real-time mapping for indoor localization systems are utilized for user navigation to the nearest exit with the added feature of floor plans for public locations that may be accessed offline for the convenience of the users. The system is also designed for instant emergency aid assistance in case of a medical personal difficulty through the use of the developed dedicated bilingual multi-platform mobile application.

Towards Secure Mobile Process Tracking Using Wearable Computing in mGovernment Applications

In this paper, we propose an automatic platform for mobile process monitoring by utilising wearable computing. The proposed system is implemented and tested for the process of the renewal of expired documents, that caters to both the clients and issuing governmental entities. The client-side system consists of a mobile application synced with a smart watch, which serves to receive notifications and status update on the renewal of the requested document. On the other hand, issuing authorities are served using specialised web portals, which offer a capability to manage the incoming requests in a structured way, to download the documents pertaining to a specific user, and to upload the processed document that is requested. Both sides of the proposed system are served using a database, which offers real-time updates and notifications. All documents are secured by using the latest encryption methods and digital signatures. The major benefit of the proposed system is the ability to monitor and go through the renewal process by simply using a user-friendly interface on the smart watch. We present the system workflow and design in the results, and discuss an application scenario that was proposed for the process of visa renewal in United Arab Emirates.

An integrated caregiver-focused mHealth framework for elderly care

In this paper, we propose an integrated caregiver-focused framework that aims to provide a health care and a fall detection service for elderly users. The proposed system looks at the responsibility of the elder-care from three different perspectives: maintenance of an accurate and updated health history, prevention of inappropriate dietary options, and detection of major fall accidents. We ensure a timely intervention by capitalizing on smart watches and their ability to notify the caregiver any time and anywhere. The integrated system provides the users with an organized medical journal that gives an insight of their medical status while being able to share it with their doctor Moreover, the system provides a food and nutrition guide that allows the users to evaluate their food intake both quantity and quality wise. Lastly, users can benefit from a fall detection service that uses the sensors available on the commercial smart watches and the cascade feed-forward neural network for classification. The experiments performed result in an accuracy of 93.33% of the proposed system in the classification of fall events.

Leaf spot area index: A nondestructive mangrove leaf spot estimation technique

Mangrove forests in the United Arab Emirates and the rest of the world hold a vital importance to the environment. For example, the Environmental Agency of Abu Dhabi has several programs in place for the preservation and protection of these forests due to their importance as breeding grounds for several sea species as well as the role they plan in preventing coastline erosion and reducing the impact of carbon emissions. Mangrove communities may exhibit defoliation, dieback, and death due to natural and man made reasons such as weather, insects and disease, nutrients, pollution, climate change, or population. A mechanism to assess the health level of the mangroves is important to help remedy the situation in its early stages. In this paper, we propose a nondestructive image processing technique to monitor and assess the health of mangrove populations by automatically estimating the ratio of the leaf spot area over the leaf area. The technique is based on processing images of mangrove leaves collected using a digital camera or a smart phone. The image of the leaf is analyzed to extract the contours of both the leaf and the spots within the leaf. Finally, we calculate what we propose as the Leaf Spot Area Index (LSAI). Estimating the index over a large and representative set of samples gives an improved estimation of the health of the mangrove ecosystem. Monitoring the index over time also helps in tracking the results of the programs in place to protect this important natural resource.