Resumen de Strategies for video sequence stabilisation
Ángela Ribeiro Seijas, Nadir Costa, Gonzalo Pajares, María Guijarro Mata-García
This paper describes a stabilisation method applied to a precisionagriculture scheme for weed control in crop fields. The method uses the centres ofthe crop rows as the characteristic elements to be detected and tracking acrossframes to compensate for the lateral movement of the camera. The inverseperspective transform is used to obtain a bird’s-eye view of the original imagerecorded at the field. The transformation allows us to keep the centres of the threecentral crop rows centred on the screen for further processing. The proposedalgorithm has been tested on image sequences of different fields and recorded atdifferent times and under different lighting conditions. Lateral displacements of upto 66% of the inter-row spacing have been suppressed.
