Motion detectors in the locust visual system: From biology to robot sensors
- 15 February 2002
- journal article
- review article
- Published by Wiley in Microscopy Research and Technique
- Vol. 56 (4) , 256-269
Abstract
Motion detectors in the locust optic lobe and brain fall into two categories: neurones that respond selectively to approaching vs. receding objects and neurones that respond selectively to a particular pattern of image motion over a substantial part of the eye, generated by the locust's own movements through its environment. Neurones from the two categories can be differentiated on the basis of their response to motion at a constant velocity at a fixed distance from the locust: neurones of the first category respond equally well to motion in any direction whereas neurones in the second category respond selectively to one preferred direction of motion. Several of the motion detectors of the first category, responding to approaching objects, share the same input organisation, suggesting that it is important in generating a tuning for approaching objects. Anatomical, physiological, and modelling studies have revealed how the selectivity of the response is generated. The selectivity arises as a result of a critical race between excitation, generated when image edges move out over the eye and delayed inhibition, generated by the same edge movements. For excitation to build up, the velocity and extent of edge motion over the eye must increase rapidly. The ultrastructure of the afferent inputs onto the dendrites of collision sensitive neurones reveals a possible substrate for the interaction between excitation and inhibition. This interpretation is supported by both physiological and immunocytochemical evidence. The input organisation of these neurones has been incorporated into the control structure of a small mobile robot, which successfully avoids collisions with looming objects. The ecological role of motion detectors of the second category that respond to image motion over a substantial part of the visual field, is discussed as is the input organisation that generates this selective response. The broad tuning of these neurones, particularly at low velocities (<0.02°/s), suggests they may have a role in navigation during migratory flights at altitude. By contrast, their optimum tuning to high‐image velocities suggests these motion detectors are adapted for use in a fast flying insect, which does not spend significant time hovering. Microsc. Res. Tech. 56:256–269, 2002.Keywords
This publication has 68 references indexed in Scilit:
- Collision avoidance using a model of the locust LGMD neuronRobotics and Autonomous Systems, 2000
- Insect motion detectors matched to visual ecologyNature, 1996
- Retinal image size triggers obstacle avoidance in flying locustsThe Science of Nature, 1993
- Evidence for two distinct movement-detecting mechanisms in insect visionThe Science of Nature, 1993
- From insect vision to robot visionPhilosophical Transactions of the Royal Society B: Biological Sciences, 1992
- Mechanisms of early visual processing in the medulla of the locust optic lobe: How self-inhibition, spatial-pooling, and signal rectification contribute to the properties of transient cellsVisual Neuroscience, 1991
- Depth perception by motion parallax and paradoxical parallax in the locustThe Science of Nature, 1990
- Course correction circuitry translates feature detection into behavioural action in locustsNature, 1985
- The separation of visual axes in apposition compound eyesPhilosophical Transactions of the Royal Society of London. B, Biological Sciences, 1978
- Protection from habituation by lateral inhibitionNature, 1975