Assisted cutting of digital images and videos

• Grant number: 97/13306-6
• Support Opportunities: Research Grants - Innovative Research in Small Business - PIPE
• Start date: June 01, 1998
• End date: September 30, 2001
• Field of knowledge: Physical Sciences and Mathematics - Computer Science - Computing Methodologies and Techniques
• Principal Investigator: Robert Liang Koo
• Grantee: Robert Liang Koo
• Company: SDC Engenharia de Sistemas Ltda
• City: São Paulo

Abstract

The SDC Company has been developing, through the project Softex, a "tool box" of mathematical morphology for the software of the visual program Matlab. Mathematical morphology is a non-linear technique for handling digital images for image segmentation. This proposal intends to develop cutting tools of the type that can easily be transferred for use in the different programs already in use on the market, such as Photoshop, Matlab itself, 3DStudio, etc. The manipulation of a photograph and of a video demands various sophisticated techniques of image processing. The growing capacity for storage and transmission of current data is possible thanks to technologies such DVD, TV digital, Internet and the various techniques of compressing data (MPEG2 and MPEG4). This ease of access to images and videos is allowing video signals to be manipulated with much more complexity within the digitalized computer. This new panorama opens up routes for diverse applications of image processing where the cutting of the objects is one of the most fundamental operations. The cutting consists in using the computer to define, on the image, automatic cuts, around objects of interest. The intention is to study the best methodology for the development of this software. In particular, during phase 1 of the research we will be looking to understand what are the users’ real needs; what sort of interface there should be; whether a stand-alone product needs to be generated or plug-ins for platforms already in existence; and which are the most promising tools. (AU)