The presence of a spatial memory deficit in Parkinson’s disease (PD) is still a matter of discussion. Nineteen PD patients and 16 controls were given two spatial tests and a non-spatial task. First, the subject was led into a room containing 4 objects and had 10 s to memorize their location. After being led outside, the subject had to place icons representing the objects on a map of the room. Differences between the real and estimated locations were evaluated. Afterwards, the subject had to choose a map showing the correct arrangement of objects from 4 alternatives. Locations of some objects were changed before the second test. The subject had 10 s to detect these changes. One point was given for each change or its absence detected. In the non-spatial working memory task, 8 cards of different shapes were used. The subject had to select a different card each time while the cards were shuffled between choices. Errors consisted of selecting previously chosen cards. The means of the above measures for both groups were compared. Absence of any significant differences suggests that PD patients perform well in "real life" memory tests in contrast to similar computerized tests.
This book introduces the furidaniental concepts and tools involved in the design and implementation of object recognition systems. Divided into three parts, it first introduces the topic and covers the acquisition of images, then details of the 3-D object reconstruction, modeling and matching, and finally describes typical recognition systems using case studies. Key features include: Extensive literature surveys of state-of-the-art systems and FTP site from which readers can obtain MATLAB codes used to generate some of the results found in the text. Recognition will be essential reading for research scientists, advanced undergraduate and postgraduate students in computer vision, image processing and pattern classification. It will also be of interest to practitioners working in the field of computer vision.