We propose the use of Difference Histogram method for classification of soya seed speckle images. The time history speckle patterns (THSP), obtained from the dynamic speckle patterns of the seeds were processed for texture classification based on seed vigor and viability. In this work, bean seeds were analyzed with different humidity levels, ie. at 15, 25, 35 and 45 minutes since the sample was taken out from the humid germination paper and submitted to the imaging process, with the aim of determining the influence of this temporal parameter in the classification result (dead or alive). The whole set of seeds was previously analyzed and classified in viable (alive) or not viable (dead) by experts in the matter by applying a traditional method. According to the obtained results the proposed method revealed to be appropriate for the task of classification of seeds. In the case of the highest humidity level the disagreement between our method and the conventional one was the greatest. It should be said that in this case the analyzed images were noisier.
Notas
Lecture Notes in Computer Science (LNCS, vol. 2905)
Información general
Fecha de exposición:2003
Fecha de publicación:2003
Idioma del documento:Inglés
Evento:8th Iberoamerican Congress on Pattern Recognition (CIARP 2003) (Havana, Cuba, November 26-29, 2003)
Institución de origen:UID en Óptica, Procesamiento de Imágenes y Metrología Óptica; Comisión de Investigaciones Científicas de la provincia de Buenos Aires
Excepto donde se diga explícitamente, este item se publica bajo la siguiente licencia Creative Commons Attribution-NonCommercial-ShareAlike 4.0 International (CC BY-NC-SA 4.0)