FLORENCE, SC, May 01, 2007—ACS Technologies is pleased to announce that employee Michael Dropps was selected by a panel of experts for publication in the academic periodical International Journal of Computer Vision.
Dropps, a software developer for the nation’s leading church management solution, ACS, co-authored the paper with two other students and a professor at the University of South Carolina while working on his Bachelor’s degree.
The printing of the Dropps’s research was great news to ACS Technologies, who exercises programs to promote excellence among employees. Steve Cumbia, executive director of research and development at ACS Technologies said, “The fact that Michael is credited with publishing a paper in an academic journal indicates that we have a very high standard for hiring our developers.” He added, “We hire excellence, and Michael is an example of that.”
The paper, entitled “Global Detection of Salient Convex Boundaries,” documents their computer vision research and the software they developed to detect the most prominent object in an image. The program they created is capable of detecting convex objects in an image and ranking them from most prominent to least prominent.
Dropps and another undergraduate student, Adam Bailey, wrote the initial steps for the software’s code in C++ using Linux workstations. Their code analyzed and reconstructed the edges of images using Bezier curves, which decreased processing time and allowed the detected objects to consist only of accurate, smooth boundaries. This step also helped to counteract noise in the images.
Dropps and Bailey’s results were further analyzed by code written by graduate student Joachim Stahl, which ensured that the resulting boundaries were convex and traversed certain points in the image.
The processed results were then run through the Ratio Contour edge-grouping algorithm written by Dr. Song Wang, USC professor. This algorithm produced the final closed boundaries within the image.
The group tested their work by creating a database of images that each had one clear focal point, selecting the obvious boundaries by hand, and then ensuring that the program detected the same boundary on its own. The group tested a total of about 1,000 images using this method.
Dropps and his teammates documented their results in a 23-page paper with the intent to be published. Dropps said, “Publication is very important because it allows other researchers to have a base on which to work and expand upon.”
The group submitted the paper to a team of computer vision professionals, most of whom are published authors. These individuals looked for clear presentation of mathematical proofs and content.
The journal selected the paper to be published in March of 2007. The paper had previously been published online in June of 2006, but its publication in International Journal of Computer Vision marked its first time in print.
Dropps added, “I encourage other undergraduate students to talk to their professors and get involved outside of class. This experience was invaluable to me, and I look forward to seeing how others will build upon our work in the future.”
The publication is available on the publisher’s Web site at www.springerlink.com.
ACS Technologies Group, Inc.
ACS Technologies (www.acstechnologies.com) is the leading provider of information management solutions for churches, schools and other faith-based organizations and has made its most notable contribution to the church management software industry with its powerful desktop application, ACS™. By meeting client needs with a variety of desktop and Web-based software solutions, customer support and professional services, ACS Technologies offers the broadest range of products and services to the largest client base in the industry—over 22,000 different organizations rely on ACS Technologies for their daily operations. The company has developed a robust product line that offers total integrated solutions for clients, providing tools to manage finances, membership, school data, church growth, Web sites, Catholic organizations through the PDS™ product line and more. ACS Technologies’ mission is simple, our solutions are focused - we empower our clients with specially designed software solutions to meet their everyday needs. Founded in 1978, ACS Technologies is a privately held company headquartered in Florence, South Carolina with over 300 employees and sales and training offices throughout the U.S.
