Accession Number : ADA292789

Title :   Visual Processing of Object Velocity and Acceleration.

Descriptive Note : Final technical rept. 15 Jan 92-14 Jan 95,

Corporate Author : SMITH-KETTLEWELL EYE RESEARCH INST SAN FRANCISCO CA

Personal Author(s) : McKee, Suzanne

PDF Url : ADA292789

Report Date : 06 MAR 1995

Pagination or Media Count : 17

Abstract : The main objective of this project was the analysis of the detection of trajectory signals in random motion noise. Results from experiments demonstrated that detection was limited by the probability of mismatch, not noise density per se; detection was based on a motion signal, not a form or position signal; detection was limited by higher-order aspects of motion processing, not the ratio of signal to noise in the primary motion detectors. The Yuille-Grzywacz "Motion Coherence" model was modified to provide a computational model for these findings. Computer simulations showed that this model was adequate to reproduce our results. In addition, three other projects were supported by this grant: motion transparency from speed signals alone; speed discrimination for targets moving in depth; trajectory motion detection in three-dimensional noise. Nine papers from this project have been accepted in referred journals; one additional paper has been submitted. Four chapters were also supported by this grant. (AN)

Descriptors :   *IMAGE PROCESSING, *VISUAL PERCEPTION, VELOCITY, MATHEMATICAL MODELS, SIGNAL PROCESSING, COMPUTERIZED SIMULATION, STIMULATION(PHYSIOLOGY), EXPERIMENTAL DATA, PERFORMANCE(HUMAN), ACCELERATION, PROBABILITY, SIGNAL TO NOISE RATIO, TARGET DISCRIMINATION, THREE DIMENSIONAL, TARGET DETECTION, BACKGROUND NOISE, VISUAL ACUITY, IMAGE MOTION COMPENSATION, TRANSPARENCIES.

Subject Categories : Anatomy and Physiology

Distribution Statement : APPROVED FOR PUBLIC RELEASE