Light-based methods are being further developed to meet the growing demands for food in the agricultural industry. Optical imaging is a rapid, non-destructive, and accurate technology that can produce consistent measurements of product quality compared to conventional techniques. In this research, a novel approach for seed quality prediction is presented. In the proposed approach two advanced optical imaging techniques based on chlorophyll fluorescence and chemometric-based multispectral imaging were employed. The chemometrics encompassed principal component analysis (PCA) and quadratic discrimination analysis (QDA). Among plants that are relevant as both crops and scientific models, tomato, and carrot were selected for the experiment. We compared the optical imaging techniques to the traditional analytical methods used for quality characterization of commercial seedlots. Results showed that chlorophyll fluorescence-based technology is feasible to discriminate cultivars and to identify seedlots with lower physiological potential. The exploratory analysis of multispectral imaging data using a non-supervised approach (two-component PCA) allowed the characterization of differences between carrot cultivars, but not for tomato cultivars. A Random Forest (RF) classifier based on Gini importance was applied to multispectral data and it revealed the most meaningful bandwidths from 19 wavelengths for seed quality characterization. In order to validate the RF model, we selected the five most important wavelengths to be applied in a QDA-based model, and the model reached high accuracy to classify lots with high-and low-vigor seeds, with a correct classification from 86 to 95% in tomato and from 88 to 97% in carrot for validation set. Further analysis showed that low quality seeds resulted in seedlings with altered photosynthetic capacity and chlorophyll content. In conclusion, both chlorophyll fluorescence and chemometrics-based multispectral imaging can be applied as reliable proxies of the physiological potential in tomato and carrot seeds. From the practical point of view, such techniques/methodologies can be potentially used for screening low quality seeds in food and agricultural industries.

Determination of physical, chemical and biological attributes with individual analyses is inadequate for improving the understanding of soil conditions as a function of land-use change (LUC) in comparison to the natural state of soil. For a more accurate soil condition diagnostic, it is necessary to consider various indicators related to these characteristics, which requires the use of multivariate statistical analysis. The aim of this work was to characterize, through multivariate analysis, different types of LUCs in an Oxisol as a function of the physical, chemical and biological attributes and to clarify the relationship of these attributes with the quality of the soil in comparison to these attributes in natural soil conditions, in the southern Amazon in Brazil. The land uses evaluated in the municipality of Alta Floresta, state of Mato Grosso (MT), Brazil, were native amazon forest (ma), degraded pasture (pd), managed renewed pasture (pn), permanent preservation area in recovery (app), crop area (rice), forage sugarcane (ca) and reforested area with eucalyptus (eu). To characterize the physical and chemical soil attributes, samples were collected in each land-use area, at depths of 0-0.10 and 0.10-0.20 m, and the determination of soil microbial activity (biological attributes) was evaluated at a depth of 0-0.10 m. The interrelationship between the analyzed attributes was described by multivariate techniques, which included hierarchical and non-hierarchical cluster analyses, principal component analysis, canonical correlation, and structural equation modeling. The multivariate approach for the analysis of soil attribute data was efficient in the identification of anthropogenic actions on areas in comparison to natural conditions. Together, the cluster analysis and principal components analysis identified two groups that differed mainly in terms of anthropic operations of soil tillage and liming. The land use that was most similar to the natural condition was degraded pasture, which was mainly due to K and H + Al contents, soil microporosity and soil basal respiration. Structural equation modeling indicated that the latent factor soil chemical attributes had three times greater interference (-0.5828) than the latent factor soil physical attributes (0.1735) on the latent factor soil biological attributes. Therefore, anthropic actions, especially the liming, modified soil acidity conditions, affecting the microorganisms of its flora and changing the native fungal community of the soil that was evaluated. large areas, mainly through the inadequate conversion of natural environments into agricultural areas (Fonseca et al., 2007; Rojas et al., 2016). The main impact of LUC is on in the soils, which are directly responsible for the sustainability and productivity of natural and agricultural ecosystems (Castilho et al., 2016; Novak et al., 2017; Sanabria et al., 2016). Studying the physical, chemical and biological attributes of soil in different applications and comparing these attributes to those in areas

Radiographic and multispectral image analysis have potential to be efficient, objective methods for assessing seed quality and internal insect infestation. The aim of this study was to verify the efficiency of radiographic and multispectral analysis in detecting signs and damage caused by Angoumois grain moth [Sitotroga cerealella (Olivier)] and its different developmental stages in wheat (Triticum aestivum L.) seeds. The experiment was conducted in a completely randomized design with six replications of 50 seeds. The samples were subjected to laboratory-induced infestation and after 5 and 10 d, radiographic and multispectral analysis were conducted. Afterwards, the seeds were immersed in water for 24 h and then sectioned with a cutting blade. The number of seeds with signs of eggs or oviposition, larvae, pupae, adult insects and insect galleries was quantified. The generalized linear models (GLM) methodology was used and the Tukey test (p < .05) was used to compare the means. In general, the radiographic (with or without contrast) and multispectral methods are viable tools to evaluate insect-infested and uninfested wheat seeds. Multispectral analysis was efficient only in identifying eggs on the seed surface and does not detect the presence of larvae and pupae inside the seeds.

No abstract

Agradeço a Deus pelas providências divinas para a conclusão desse trabalho. Aos meus pais, Antônio e Maria Luiza, e minha irmã Caren, que me deram suporte e apoio nas decisões que precisei tomar ao longo desse árduo caminho. Foram minha força, meu suporte, me deram colo e sempre me lembraram que se orgulham de mim, me ajudando a acreditar que o impossível era possível.Ao meu noivo Felipe que compartilhou de todas as minhas angústias e me pegou pela mão, com sua calma e generosidade, nunca me deixou desistir ou duvidou de minha capacidade e força. Tenho muita admiração por você.Ao meu orientador, Professor Carlos Tadeu, que confiou e acreditou em mim até o último instante.Aos amigos que fiz ao longo do caminho que me auxiliaram nessa etapa tão intensa e desafiadora, seja nas atribuições do trabalho ou em suporte psicológico. Em especial ao meu amigo Welinton que sempre se prontificou a me ajudar.À @Tech por me ceder os dados para a realização deste trabalho.Aos professores e aos funcionários do Departamento de Ciências Exatas da Esalq. Aos membros da banca de qualificação e defesa que gentilmente contribuíram para a melhoria do trabalho.O presente trabalho foi realizado com apoio da Coordenação de Aperfeiçoamento de Pessoal de Nível Superior -Brasil (CAPES) -Código de Financiamento 001, e ao Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPQ).