Viabilidade do uso de raios X no beneficiamento de sementes de soja

Abstract

Image analysis is a technique that shows potential to be used in association with the usual tests for seed evaluation, providing improvement of methodologies and obtaining reliable results. The acquisition of images by X-rays has shown to be promising for the evaluation of physical characteristics related to the physiological potential because it is a fast and non-destructive procedure, making possible the subsequent use of radiographed seeds. The X-ray images can be associated with software and the information produced allows the analysis of characteristics such as size and shape of the seed, and these same characteristics are used to select lots of soybean seeds in the beneficiation process. Thus, the present research aimed to evaluate the efficiency of the use of X-rays in determining the quality of soybean seeds in processing, in order to relate it to the physiological quality of soybean seeds. Seed samples were collected at the Uniggel Sementes processing unit, in Chapadão do Céu -GO, at two strategic points, at the beginning of the processing, when the seeds were leaving the dryer, and at the end of the process, after being bagged. The design used was completely randomized, with four treatments formed by the cultivars HO Aporé IPRO, DM 73I75 RSF IPRO, Brasmax Olimpo IPRO and Brasmax Bônus IPRO and four replications. To determine the quality of the seeds at each sampled point, the determination of water content and the tetrazolium test were performed. To obtain the images, the soybean seeds were placed on an acrylic plate containing 50 units in four replications per treatment and submitted to X-rays. Subsequently, the same radiographed seeds were submitted to the germination test, maintaining their positions. The radiographed images were processed in ImageJ® software, obtaining measurements of area, perimeter, width, length, median gray, integrated density, aspect ratio and solidity. Data were submitted to Tukey's test (p<0.05) and principal component analysis (PCA) was applied. The results obtained showed that the analysis of radiographed images of the seeds at the two collection points did not make it possible to separate the cultivars in terms of physiological quality. Thus, due to the high quality of the soybean seeds received at the processing unit, it is not possible to verify the efficiency of the analysis of X-ray images in the processing stages, due to the low correlation with the physiological quality.