Research Focus

My research interest includes Personal Authentication using Multimodal Biometrics, Inspection of Surface Defects for Quality Assurance using Automated Visual Inspection and Palmprint Verification for Automated Personal Security. For more information, have a look at my Projects.

Personal Authentication using Multimodal Biometrics

The technology for trusted e-security is critical to many business and administrative process. There has been a newfound urgency after September 11 attacks to develop cutting-edge security technologies. However, the performance of currently available technology is yet to mature for its broad deployment in real environments The fingerprint and hand geometry based biometric systems account for more than 60% of current market share. However, the security evaluation against attacks using fake fingerprint or hand has been rarely disclosed. The recent risk analysis using artificial fingers made of cheap and readily available gelatin, have shown extremely high acceptance on available fingerprint systems. The multimodal biometrics system allows integration of two or more biometric in order to cope up with the stringent performance requirements imposed for high security access. Such systems offer high reliability due to the presence of multiple piece of evidence and are vital for fraudulent technologies as it is more difficult to simultaneously forge multiple biometric characteristics than to forge a single biometric characteristic.

More on this Research:

Inspection of Surface Defects for Quality Assurance using Automated Visual Inspection

Automated inspection of surface defects is key for the quality assurance in several industrial products. Some of the most challenging visual inspection problems deal with the textured materials. In most cases, the quality inspection through visual inspection is still carried by humans. However, the reliability of manual inspection is limited by ensuing fatigue and inattentiveness. In textile industry, even the most highly trained inspectors can only detect about 70% of defects. In this research, several new algorithms were developed for the detection of local textured defects in industrial materials using various approaches. Real fabric samples obtained from the textile industry were used as prototype of textured materials. The range of problems considered included the realtime implementation of developed algorithm(s), automated selection of parameters, sensitivity of online algorithm to the yarn impurities and low-cost implementation. The developed algorithms have shown to offer high detection of defects and even most difficult textured defects were successfully detected.

More on this Research:

Palmprint Verification for Automated Personal Security

Palmprints can be used to substitute fingerprints, for personal identification, in situations arising due to failure of fingerprints. Fingerprint identification is widely used in personal identification as it works well in most cases. However it is difficult to acquire unique fingerprint features i.e. minutiae for some class of persons such as manual laborers, elderly people, etc.. Therefore some researchers have recently investigated the utility of palmprint features for the personal recognition. Prior work on palmprint identification focused on the principal palmprint features i.e. creases and lines. However, the information content of palmprint images also consists of certain local and global features, which can be used for recognition. Thus a palmprint image can be analyzed as texture, which is random rather than uniform. This texture is composed of wrinkles, ridges and principal lines, and is quite unique in every person.

The proposed technique employed inkless palmprint images that were acquired from a simple setup that does not employ pegs to cause user inconvenience. The proposed algorithms were tested on relatively small database (40 users) as the large database was not available.

More on this Research: