https://repositorio.ifgoiano.edu.br/handle/prefix/217 2026-03-24T13:23:26Z 2026-03-24T13:23:26Z Almeida, Guilherme Pires Silva de https://repositorio.ifgoiano.edu.br/handle/prefix/6137 2026-02-02T12:53:30Z 2025-11-29T00:00:00Z

Título: SISTEMAS INTELIGENTES EMBARCADOS EM DISPOSITIVOS MOVEIS E BASEADOS EM DEEP LEARNING PARA DETECÇÃO E MONITORAMENTO ESPACIAL DE PRAGAS AGRÍCOLAS Autor(es): Almeida, Guilherme Pires Silva de Primeiro Orientador: Santos, Leonardo Nazário Silva dos Primeiro Membro da Banca: Teixeira, Marconi Batista Segundo Membro da Banca: Oliveira, Mario Anderson de Terceiro Membro da Banca: Morais, Wilker Alves Abstract: Accurate and timely detection of insect pests remains one of the major challenges in modern agriculture, especially in large-scale soybean and maize production systems. Inefficient monitoring practices often result in delayed control interventions and significant yield losses. Recent advancements in deep learning and mobile computing have opened new opportunities for in-field pest identification using lightweight computer vision models. In this context, this thesis presents an integrated framework for intelligent pest detection and spatial monitoring based on deep learning, geostatistical analysis, and mobile applications. First, two datasets of insect pests were constructed and evaluated: a comprehensive high-resolution dataset curated through double-expert validation, and a smaller sample designed for comparative analysis. State-of-the-art detection architectures (YOLO and Detectron2) were trained on both datasets and subsequently converted into TensorFlow Lite (TFLite) and ONNX formats to enable deployment on resource-constrained devices. Even under the least favorable conditions using the reduced dataset and the lightest ONNX model the results reached a precision of 87.3% and accuracy 95.0%, demonstrating the robustness of the pipeline. Building upon these results, a mobile system named AgroInsect was developed. The application performs real-time, on-device detection of four key pest species relevant to Brazilian soybean and maize production (Diabrotica speciosa, Dalbulus maidis, Diceraeus spp., and Spodoptera frugiperda), automatically extracts geolocation metadata, validates spatial consistency based on field boundaries, and synchronizes detections with a cloud database. Spatial visualization is generated through heatmaps and Ordinary Kriging (PyKrige), enabling high-resolution incidence maps. Field evaluations confirmed strong model performance, with overall accuracy of 95.1%, F1-scores above 0.94 for all species, and only 1.1% false detections. The kriging model achieved R² > 0.94 under dense sampling, accurately reproducing ecological spatial patterns. Additionally, this thesis introduces AgroLabIA, a digital platform designed for the storage, annotation, and dissemination of agricultural pest datasets. It provides curated, multi-format datasets suitable for training machine learning models and supports the continuous expansion of new insect and weed classes. The integrated environment that encompasses dataset generation, mobile detection, spatial verification, and geostatistical mapping demonstrate a scalable and operationally robust solution for precision pest monitoring. The results position the AgroInsect database as an effective tool for accelerating decision-making in integrated pest management, particularly in regions with limited connectivity, thus contributing to the consolidation of Agriculture 4.0. Editor: Instituto Federal Goiano Tipo: Tese

2025-11-29T00:00:00Z Lora, Jacyr https://repositorio.ifgoiano.edu.br/handle/prefix/6119 2026-01-15T13:19:01Z 2025-06-23T00:00:00Z

Título: EFEITO RESPONSIVO DA CULTURA DO MILHO PELA APLICAÇÃO DE COMPOSTOS MINERAIS Autor(es): Lora, Jacyr Primeiro Orientador: Soares, Frederico Antonio Loureiro Primeiro Membro da Banca: Cunha, Fernando Nobre Segundo Membro da Banca: Silva, Nelmicio Furtado da Terceiro Membro da Banca: Morais, Wilker Alves Abstract: The use of mineral compounds in agriculture has emerged as a promising strategy to increase crop productivity and sustainability. Many mineral compounds, although not considered fertilizers, stimulate natural processes by providing essential nutrients that participate in various physiological and growth processes, and even improve water use efficiency, tolerance to abiotic stresses, and crop quality. This thesis aimed to evaluate the effects of silicon and zinc fertilization on growth, development, and yield of corn (Zea mays L., grown in Cerrado Red Latosol). Two experiments were carried out in open-air pots using a randomized block design with five doses (0, 25, 50, 100, and 200 kg/ha⁻¹) of potassium silicate and zinc sulfate, respectively, and four replicates. The variables evaluated included morphological, physiological, and productive parameters namely, plant height, stem diameter, number of leaves, leaf length, leaf width, leaf area, leaf mass, stem mass, SPAD index, NDVI index, chlorophyll a, chlorophyll b, total chlorophyll and chlorophyll a/b. The productive components evaluated were straw mass, ear mass, cob mass, ear length, ear diameter, cob diameter, number of rows of grains in the ear, number of grains per row and the mass of 50 grains. In the first experiment, the silicon aplication provided significant increases in plant height, stem diameter, leaf area, chlorophyll content (SPAD and NDVI indices), mass of vegetative organs and ear components, with emphasis on the maximum productivity of 5582.44 kg ha⁻¹ at the dose of 118.31%. The response was predominantly quadratic, indicating an optimum application point for most variables. In the second experiment, zinc also positively influenced the variables studied, especially at intermediate doses. The best results occurred between 100% and 140% of the standard dose, with maximum productivity of 4946.30 kg ha⁻¹ at the dose of 135.56%. Both mineral compounds demonstrated potential to improve the agronomic performance of corn, with silicon being more effective in structural attributes and zinc being more associated with enzymatic metabolism and leaf pigmentation. The data reinforces the importance of rational nutritional management and the need to define adequate application ranges to maximize positive effects and avoid losses due to excess. Editor: Instituto Federal Goiano Tipo: Tese

2025-06-23T00:00:00Z Lourini, Simone Hemkemeier https://repositorio.ifgoiano.edu.br/handle/prefix/5944 2025-12-05T17:22:35Z 2025-10-08T00:00:00Z

Título: MANEJO DE ADUBAÇÃO FOSFATADA COM FERTILIZANTES DE EFICIÊNCIA AUMENTADA Autor(es): Lourini, Simone Hemkemeier Primeiro Orientador: Rubio Neto, Aurélio Primeiro Membro da Banca: Rubio Neto, Aurélio Segundo Membro da Banca: Rodrigues, Carlos Ribeiro Terceiro Membro da Banca: Menezes, June Faria Scherrer Abstract: Phosphorus (P) is vital for agricultural productivity and global food security. Although global reserves of phosphate rock remain available, their geographically concentrated distribution requires more efficient nutrient use and reduced external dependence. In Brazilian tropical soils, high acidity and abundance of iron and aluminum oxides intensify P fixation and lower its agronomic efficiency, particularly when highly soluble sources are used. Smart phosphorus management strategies, guided by the 4R principles (right source, right rate, right place, and right time), have gained attention for combining efficiency, sustainability, and circular-economy approaches. In this context, organic and polymeric fertilizer additives represent promising alternatives to modulate P release and minimize losses. To evaluate these effects, greenhouse incubation experiments, structural equation modeling (SEM), and field trials with soybean and maize were carried out to assess P dynamics and utilization under contrasting soils and management conditions. Three soils were tested: Acriferric Red-Yellow Latosol (LVAwf), Typic Orthic Quartzarenic Neosol (RQo), and Typic Eutrophic Red Nitosol (NVe). Each soil followed a 2 × 3 + 1 factorial design with two P sources (MAP and ST), three additive conditions (none, KimcoatBioHP®, and IFCoat), and one control, with soil samples collected at 30, 90, 180, 270, and 360 days. Phosphorus fractions were determined by the Hedley sequential fractionation method (Hedley et al., 1982), in which soluble P was extracted using ammonium chloride (NH₄Cl). Results were integrated using multivariate analyses and SEM, treating measured fractions as observed variables associated with latent constructs (Organic, Inorganic, Mineral, Occluded, and Available Pools), allowing the identification of the main determinants of P availability in each soil. Phosphorus dynamics varied across soils, sources, and additives. In the RQo, a sandy soil with low cation-exchange capacity, MAP combined with additives increased soluble Pi at 30 days, but P was rapidly converted to non-labile (Pi-NL) and occluded (Pi-occ) forms, reducing its persistence. From 90 days onward, ST without additives sustained higher availability, indicating that slow-release sources and split applications are more suitable for this soil. In the LVAwf, rich in Fe and Al oxides, strong initial immobilization occurred, with occluded P acting as a secondary reserve. MAP + IFCoat reduced precipitation into insoluble forms and maintained higher soluble and labile Pi between 90 and 180 days, while ST showed slower dissolution and greater Ca-phosphate retention. In the NVe, with higher clay and organic-matter content, microbial cycling influenced P transformation, increasing microbial (Po-mic) and moderately labile organic (Po-ML) fractions, which acted as transient reservoirs. SEM results confirmed distinct controlling pools: the inorganic pool in RQo, occluded pool in LVAwf, and organic pool in NVe, showing that soil mineralogy and biology govern P maintenance in solution. Field experiments evaluated source and additive effects on P availability and crop performance. Soybean was grown under a completely randomized design with MAP applied in rows or broadcast, combined with six PhytohumusBio® doses (0–7.5 kg Mg⁻¹). Maize was grown in randomized blocks (2 × 4 + 1) with two sources (MAP and MAP + KimcoatBioHP®), four P rates (70, 85, 100, and 115 % of the recommendation), and a control. In soybean, PhytohumusBio® increased grain yield by 550 kg ha⁻¹ and P agronomic efficiency by 73.2 kg grain kg⁻¹ additive compared with MAP alone, especially under row application. For maize, MAP + KimcoatBioHP® reduced early growth but did not affect final yield, due to climatic stress during the season. Overall, the results demonstrate that humic and polymeric additives associated with phosphate fertilizers are promising tools to improve P use efficiency. Their effectiveness, however, depends on soil mineralogy, biological activity, and management strategy. The smart phosphorus management concept, integrating the right soil, source, and additive, proved essential to maintain P in labile forms, reduce losses, and support sustainable agriculture in tropical environments. Editor: Instituto Federal Goiano Tipo: Tese

2025-10-08T00:00:00Z Martins, Bruna Silva https://repositorio.ifgoiano.edu.br/handle/prefix/5888 2025-11-11T21:27:32Z 2025-08-29T00:00:00Z

Título: RESPOSTAS DE Licania tomentosa E Syzygium malaccense À APLICAÇÃO DE GLIFOSATO Autor(es): Martins, Bruna Silva Primeiro Orientador: Carlos, Leandro Abstract: The agriculture intensification driven by the agribusiness expansion in Brazil has led to the increasing use of herbicides such as glyphosate. Although effective in weed control, its extensive application has raised concerns about side effects on non-target plant species, especially those located near cultivated areas. In this context, the present study aimed to evaluate the physiological and morpho-anatomical effects of glyphosate on two native tree species widely used in urban afforestation and found in Cerrado ecosystems: Licania tomentosa and Syzygium malaccense, investigating their sensitivity to herbicide and their potential as environmental bioindicators. The experiment was carried out in a greenhouse using healthy seedlings of both species exposed to five increasing doses of glyphosate (0, 25, 50, 75, and 100 g ha⁻¹) in a completely randomized design. The evaluations included biometric analyses (plant height and stem diameter), physiological assessments (gas exchange), and morpho-anatomical analyses of the leaves. For Licania tomentosa, the results showed no visible symptoms of phytotoxicity, but significant internal alterations such as necrosis in idioblasts, loss of chlorophyll parenchyma, and epidermal hyperplasia. Physiologically, reductions were observed in transpiration rate, CO₂ assimilation, and stomatal conductance, along with an increase in internal CO₂ concentration and a decrease in carboxylation efficiency, indicating severe metabolic stress. However, growth variables did not show statistically significant changes. In Syzygium malaccense, no visible phytotoxic symptoms were observed either, but progressive anatomical changes were detected, such as epidermal duplication and necrosis, along with the disappearance of idioblasts containing calcium oxalate crystals. There was a significant reduction in plant height and stem diameter, as well as a negative impact on gas exchange, reflecting impaired photosynthesis. The increase in internal CO₂ suggests failures in photosynthetic biochemical processes, even at reduced herbicide doses. Both species demonstrated sensitivity to glyphosate, evidenced by physiological and structural changes prior to the appearance of visual symptoms. These characteristics highlight the potential of L. tomentosa and S. malaccense as bioindicators of environmental contamination in transitional areas between native vegetation and agricultural fields. The early detection of internal damage reinforces the importance of anatomical and physiological methods in assessing the effects of herbicides on non-target plants, supporting monitoring and conservation strategies. Editor: Instituto Federal Goiano Tipo: Tese

2025-08-29T00:00:00Z