Warning:
This page is
provided for historical and archival purposes
only. While the seminar dates are correct, we offer no
guarantee of informational accuracy or link
validity. Contact information for the speakers, hosts and
seminar committee are certainly out of date.
The talk presents pseudo-random encoding as an efficient alternative for 1-D, 2-D, and 3-D absolute position recovery with applications to robot sensing and object recognition.
Absolute position recovery paradigm using pseudo-random encoding can be stated as follows: "Given a pseudo-random encoded object-surface, whose grid node coordinates are defined in a 3-D reference frame, and given a perspective image showing only a part/window of the PRBA, recover the 3-D position and orientation of the observer's frame relative to the object reference frame."
Pseudo-random encoding is based on the window properties of the "pseudo-random multi-valued sequence" (PRMVS) and "pseudo-random multi-valued array" PRMVA.
A PRMVS has multi-valued entries taken from an alphabet of q symbols, where q is a prime or a power of a prime. Such a sequence is generated by an n-position shift register with a feedback path specified by a primitive polynomial of degree n with coefficients from the binary Galois field GF(q). According to the PRMVS window property any q-valued contents observed through a n-position window sliding over the PRMVS is unique and fully identifies the current position of the window.
A PRMVA can be obtained by folding a PRMVS. The 2-D array-index recovery is based on the PRMVA window property. According to this any pattern seen through a small window sliding over the PRMVA is unique and may fully identify the window's absolute position index (i,j) within the PRMVA.
A PRMVS has multi-valued entries taken from an alphabet of q symbols, where q is a prime or a power of a prime. Such a sequence is generated by an n-position shift register with a feedback path specified by a primitive polynomial of degree n with coefficients from the binary Galois field GF(q). According to the PRMVS window property any q-valued contents observed through a n-position window sliding over the PRMVS is unique and fully identifies the current position of the window.
A PRMVA can be obtained by folding a PRMVS. The 2-D array-index recovery is based on the PRMVA window property. According to this any pattern seen through a small window sliding over the PRMVA is unique and may fully identify the window's absolute position index (i,j) within the PRMVA.
Different techniques are discussed for the translation of the pseudo-random window code into more convenient natural representation as requested by practical applications.
The talk concludes with the presentation of a number of 1-D, 2-D and 3-D applications of the pseudo-random encoding: shaft encoders, AGV position recovery, mobile robot navigation, model based object recognition (using vision or touch), and structured light.
Emil M. Petriu received the Dipl. Eng. and Dr. Eng. degrees in electrical engineering from the Polytechnic Institute "Traian Vuia," Timisoara, Romania, in 1969 and 1978, respectively. He currently is Professor and Chairman of the Department of Electrical Engineering at the University of Ottawa, in Ottawa, Ontario, Canada, where he has been since 1985. Dr. Petriu's research interests include sensors for robotics, telepresence, neural networks and fuzzy systems for instrumentation. He has co-authored
The talk concludes with the presentation of a number of 1-D, 2-D and 3-D applications of the pseudo-random encoding: shaft encoders, AGV position recovery, mobile robot navigation, model based object recognition (using vision or touch), and structured light.