The object of this proposal is to test the hypothesis that polarization vision (like color vision) has the capacity to segment the visual field based upon spatial and/or temporal variations in e-vector orientation. As such polarization contrast can provide a complete basis for form and motion vision. For maximum utility such a system should be available over all of an organism's visual field, it should operate across the entire visible spectrum, it should be independent of specific electric-vector (e-vector) angles and it should operate with a modest degree (_50 %) of polarization. Behavioral and neurophysiological studies of crayfish optomotor reflexes are proposed to examine the limits of polarization vision for the detection of self-motion. In the proposed study, moving images will be formed by polarized light and segmented only by variations in e-vector angles. The stimuli will be produced by software driving a modified LCD projection panel. Eyestalk movements are measured with an inductive transducer. Visual interneurons and the motoneurons of the optomotor pathway will be studied with conventional intracellular and extracellular recording procedures while the neurons respond to the identical stimuli used in the behavioral studies. The behavioral studies will reveal the limitations of polarized light as a source of perceptual contrast and the extent to which the system is confined by intrinsic filters, e.g. a bias toward certain evector angles. Specific measurements include the dependence or independence on absolute e-vector angles, the sensitivity to differences in e-vector angles (at segmentation boundaries), and the dependence of the system upon the degree of polarization. The neurophysiological studies will determine the strategies employed by the neural pathway to extract the relevant information. The working hypothesis is that the stimulus e-vector angles are detected by selective photoreceptors, and the signals related to the spatial e-vector distribution are refined by documented polarization-opponent interneurons in the second optic neuropile (medulla externa). An array of medullary interneurons are known to converge on optomotor neurons by direct and indirect excitatory and inhibitory synaptic pathways. In the optomotor neurons the afferent information is integrated and the behavioral command is formed by the pattern of activity across the efferent ensemble. Further insight into the coding strategies of the system will be gained from the differences between visual interneuron and motoneuron responses to polarized images. It is anticipated that the proposed study will have a broad impact in three areas. The demonstration that polarization contrast is a general basis for segmentation of the visual field will arouse the interests of a wider audience of biologists and vision scientists because it implies a fundamental role for polarized light in form vision and motion detection. A second area of potential impact is in machine vision systems and remote sensing where the integration of polarizers and photodetectors is already proceeding on the basis of prior biological research. Finally the proposed project will provide an environment for training the next generation of scientists from student participation in the project. The fundamental validity of the the hypothesis can be ascertained in about one year.
