Yield maps generated during the harvesting operation are made from a great number of georeferenced points. Each point shows the yield of a limited area around it and they all may be interpolated to generate a representation of surface. Not always the points represent the correct yield because of errors during data collecting process. The amount and type of errors depend on yield monitor and combine model, size and shape of fileds, GPS signal quality and others. Errors on yield mapping data will cause missinterpretation on field variability, reducing map quality. Errors may be eliminated by using spredshets and in this work were proposed and used logic functions for an algorithm that detect and eliminate most frequent and important errors on raw data. The algorithm was tested with data from several fields and here is presented the example of two of those fields showing the improvement of yield maps quality.
Keywords: yield monitor, filtering, GPS