Rahman Tashakkori

Leaf classification utilizing a convolutional neural network

Tree identification is an important task in biological research. Since not all biologists are qualified for this task and experts are not accessible all the time, an application to assist in plant identification would be extremely useful. This paper describes an application classifies the type of a tree based on a picture of one of its leaves. The system developed for this research utilizes a convolutional neural network in an android mobile application to classify natural images of leaves.

Early participation of CS students in research

Research experiences are widely available to upper-division computer science students during the academic year and during summer. Co-op and internship opportunities are available to this group as well. Due to the fact that freshman and sophomore students do not have sufficient background, they are often left behind and are not involved in research activities. This paper shares some experiences with a program that was put in place through an NSF STEP grant that provides research opportunities to freshman, sophomore, and first year transfer students. The paper presents examples of projects in which computer science scholars were involved. We have learned that lower-division computer science students are excited about the opportunity to participate in research. Early participation in research helps build a strong community among the freshman and sophomore students, keeps these students engaged, and results in a higher retention rate.

Automated beehive surveillance using computer vision

The number of honey bees entering and leaving the hive throughout the day is an important metric for beekeepers. Some commercial systems that utilize infrared sensors at the hives entrance exist for monitoring honeybee traffic. This research explores a solution that is based on visual information obtained through videos taken in front of the hives. This paper describes a system for surveillance based on traditional object detection and tracking methods. The project compares results from different algorithms within the system and discusses how the performance of the system can be effectively evaluated.

Radiation damage studies for the SDC electromagnetic calorimeter

Abstract We report the results from a year long study aimed at radiation resistance and optical performance of scintillator tile with green wave shifter fiber readout. A careful investigation of several rad-hard plastic scintillators from Bicron and Kuraray, studies indicate that for a specific rad-hard Bicron scintillator, it is possible to build a tile/fiber EM calorimeter that can operate in the design luminosity of SSC. This calorimeter with excellent optical response would only have a light loss of about 5% after being exposed to 1 Mrad.

Research experience for teachers: data analysis & mining, visualization, and image processing

Our department received funding from the National Science Foundation to establish a three-year Research Experience for Teachers site in Data Analysis & Mining, Visualization, and Image Processing. The objective is to provide twelve in-service high school teachers and community college faculty to work with faculty mentors and their graduate and undergraduate assistants to conduct research in these fields. During this six-week summer program, participants gain skills that they can utilize to assist their students to solve interdisciplinary problems. In addition, participants design learning modules to teach STEM concepts in their courses. The goal of our program is for teachers to bring knowledge of computer science and its application to their classroom exposing their students to computer science. This paper will share some of the activities of this experience.

Similarity analysis of voice signals using wavelets with dynamic time warping

Accurately recognizing speech is a difficult task. Differences in gender, accent, pace, tone, as well as defects in the recording equipment and environmental noise can disturb a voice signal. Speech recognition systems are commonly studied and implemented by companies trying to alleviate problems, such as illness or injury, or to increase overall efficiency. This research uses wavelet analysis with several traditional methods to study similarities among sound signals. Through a series of seven steps, a similarity analysis of some voice signals from the same speaker as well as from different speakers is performed. The efficiency of four different wavelets (Haar, db2, db4 and Discrete Morlet), different correlation methods developed previously or in this research, and two different Dynamic Time Warping methods are studied in this research. Through several experiments, it will be shown that these techniques produce excellent results for signals by the same speaker. Based on the limited number of cases studied in this research, some evidence will be presented that suggests the proposed methods on this research are more effective for recognizing male voice files than those of females.

A honeybee hive monitoring system: From surveillance cameras to Raspberry Pis

In recent years, there has been a significant loss in the number of honeybee hives due to several natural phenomena that are referred to as the Colony Collapse Disorder (CCD). Several research programs have been established to monitor honeybee hives and to understand the behavior of the bees. A reliable monitoring system is important as it provides the beekeepers an opportunity to remotely observe the progress of their hives without the need to actually open them. This paper will present the results of several years of study and investigation for creating a reliable and inexpensive hive monitoring system. Our goal has been to create an inexpensive and modular system which sends sensor data acquired from beehives to a server for further analysis. Our initial efforts utilized video surveillance system but evolved to a more reliable system that utilizes Raspberry Pis and ATmega328Ps as its framework. This paper will provide details on the initial system and how it evolved to a new system that we are calling Beemon with its data acquisition capabilities. The early video and audio recordings obtained by the system are used in some image and signal processing analysis which are not presented on this paper.

A New Mobile Online Social Network Based Location Sharing with Enhanced Privacy Protection

Location based services (LBSs), which are useful applications of mobile online social network (mOSN), exploit various geographic properties. Location sharing helps people to share their current locations with designated friends and is one important primitive to construct the LBSs. The recent reports showed that a poorly designed location sharing scheme could easily allow the privacy of users to be violated. Over years, lots of efforts are made to provide a privacy-preserving location sharing, but none of them is satisfactory. To address this issue, we introduce a new location sharing scheme in mOSNs with a strong user privacy protection mechanism such that (a) the users current location as well as (b) the list of friends who will learn the users current location will be protected from any unintended entity, while the designated friends in the list will learn the exact location of the user. For this purpose, we introduce a new cryptography primitive called the functional pseudonym scheme based on Lagrange polynomial with the public social network IDs of the designated friends. Then, the pseudonym of a user is posted on the server along with the current location of the user. While each user can see every posted messages (pseudonym and location pairs), the actual identify of the originator of each pair can be verified only by designated friends, whose identities are used to compute the pseudonym. Most importantly, unlike any of the existing counterparts, our scheme does not assume neither a trusted server nor pre-established secret among the friends.

Image processing for honey bee hive health monitoring

Honey bees are responsible for more than half of the pollination across the world, hence they have an important role in agriculture. In recent years, the number of honey bees have declined due to the Colony Collapse Disorder (CCD), which may be due to possible causes that include climate change and pollution. Since honey bees start from their hives to go for pollination, their activities around the hive are a good indicator of their health. The bee hives health status and the activities of honey bees can be related to the environment in which they live, thus finding a quantitative measure of the bee activities is very important and can benefit the beekeepers. This paper will present a data acquisition and monitoring system that utilizes image processing techniques for extracting data from the videos of the bee hives entrance taken at the front or the top of the hives entrance. The system developed as part of this research utilizes two different approaches based on the Illumination-invariant Change Detection Algorithm and Signal-to-Noise Ratio to estimate the number of bees at the entrance of the hives.

Message encoding in images using lifting schemes

Digital images store large amounts of data and information. This data can be manipulated to some extend without being detected by human eyes. An example of such manipulations is insertion of secret information which is often referred to as information hiding. A successful insertion of a message into an image is more difficult using color images than that of grayscale images. A successful information hiding should result in the extraction of the hidden data from the image with high degree of data integrity. This paper presents an information hiding technique that utilizes lifting schemes to effectively hide information in color images.