28

"Useful Fields of View (UFOV) do not cover up the whole visual field and are the part of the visual field in which it is possible to extract maximum information. UFOV are important for visual search tasks, essential in both industry and daily life. Generally, theories of visual search and mathematical models developed for the prediction of search performance have assumed it has a circular or elliptical field shape, even though previous work had shown marked irregularities in Useful Fields of View. The aim of this work was to conduct an extensive study of the individual field shape characteristics of sphericity, boundary smoothness, symmetry, elongation and regularity; and, to select representative shape indices to investigate the relationship between visual field shape and search time. Two experiments were conducted in that vein. The first mapped 24-meridian useful fields of view for 101?subjects to study the effects of UFOV shape on visual search time. It was found that UFOV shapes were elongated horizontally with a medium level of sphericity, high levels of boundary smoothness, symmetry, and regularity. Using correlations between shape indices and search time, a set of representative shape indices for sphericity, boundary smoothness, and regularity were chosen. In the second experiment, the applicability of the chosen shape indices to performance prediction on a more difficult search task was tested. It was found that the indices did enhance the accuracy of regression models. Using data from both experiments, regression models relating search time with the selected shape indices were developed. The work here suggested the usefulness of field shape indices for enhancing the accuracy of regression models of search performance. This study demonstrated that further considerations of the importance of irregularities in Useful Fields of View is necessary, and can be greatly helped through the use of the Visual Lobe Measurement System (VILOMS), developed by the authors. VILOMS makes it possible to establish a wide range of shape indices for use in models of visual search, which may provide a valuable selection and training tool for inspectors."