Análise de sementes de senna siamea lam. por meio de processamento digital de imagens

Abstract

Seed quality is one of the parameters of great importance among the criteria of installation and establishment of a species. Thus, the non-destructive and rapid determination of seed quality, such as X-ray, proves to be useful in the seed industry, especially in the decision making process, during batch separation and according to quality. Although efficient, the tools used need to be improved to improve data processing, which will strengthen test reliability and repeatability. The objective of this study was to apply a high performance morphometric methodology, using free tools in the processing of Senna siamea radiographic images, thus, relating the parameters of the physiological quality of the seeds, as well as the adequacy of methodologies for the analysis of seedling. The experiment was conducted using seeds from 5 lots of S. siamea, where ImageJ® software was used for seed and seedling measurements, correlating with traditional tests and thus evaluating the potential of physiological quality between the tests. For the traditional tests, the seeds were radiographed and then submitted to the germination test, from which the variables related to their physiology were obtained. Data were submitted to ANOVA and when significant, means were compared by Tukey test (p≤0.05). Subsequently, Pearson's simple correlation coefficients (r) were calculated between combinations of seed physiological and physical quality evaluation tests. Among the results, it was observed that the x-ray in S. siamea seeds showed efficacy in the identification of damage, showing relationship with the physiological potential. In addition, it was possible to differentiate the lots from seedling analysis, associating the relationship with the images of the radiographed seeds.