Seed quality significantly affects how well crops grow. Traditional methods for checking seed quality, like seeing how many seeds sprout or using a chemical test called tetrazolium testing, require people to look at the seeds closely, which takes a lot of time and effort. Nowadays, computer vision, a technology that helps computers see and understand images, is being used more in farming. Here, we use computer vision with X-ray imaging to assist experts in rapidly and accurately assessing seed quality. We looked at three different sets of seeds using X-ray images and used YOLOv8 to analyze them. YOLOv8 software measures different aspects about seeds, like their size and the area taken up by the part inside, called the endosperm. Based on this information, we put the seeds into four groups depending on how much endosperm they have. Our results show that the YOLOv8 program works well in identifying and separating the endosperm, even with a small amount of data. Our method was able to accurately identify the endosperm about 95.6% of the time. This means that our approach can help determine how effective the seeds are to plant crops.

O reconhecimento de espécies de abelhas e demais polinizadores pode contribuir diretamente para a conservação de ecossistemas. O objetivo desse trabalho é desenvolver um modelo de processamento digital de imagens, para reconhecimento automático de abelhas melíferas (Apis mellifera L.) entre as demais espécies de abelhas e outros insetos, por meio do uso de redes neurais convolucionais. Para treinamento do modelo proposto foi utilizado um conjunto de dados composto por 2.300 imagens, separadas entre a espécie de abelha melífera e demais insetos (incluindo outras abelhas nessa classe), as quais foram denominadas de "Não Apis". O modelo de rede neural proposto respondeu com 94% de acurácia, culminando em um modelo de predição com um alto índice de precisão capaz de reconhecer imagens de abelhas melíferas e diferenciá-la das demais abelhas e outros insetos.

Reliable seed vigor assessment is still limited by the lack of accurate and reproducible measurements of seedling structures. Although computer vision offers a scalable alternative, segmentation remains a bottleneck due to the high visual similarity between morphological parts and severe class imbalance, particularly for hypocotyls. We propose COPPER-UNet, a U-Net-based architecture designed to overcome these constraints. The model combines multi-scale context aggregation, channel attention, and residual learning to improve feature discrimination, while an ASPP module captures global context. Across multiple benchmarks over several architectures, COPPER-UNet outperformed established U-Net variants and state-of-the-art segmentation models, achieving 79.97% mIoU, 91.39% Dice, and consistent gains across all classes, including hypocotyls (70.16% IoU). The model also provides a better balance between accuracy and computational cost. COPPER-UNet reduces segmentation errors in challenging seedling structures and enables more reliable, automated extraction of morphological traits for seed vigor assessment.

Agricultural crops productivity is heavily influenced by the strength of their seeds. Current methods for testing seed vigor, such as the germination rate and tetrazolium test, rely on time-consuming human visual inspections. Computer Vision, a crucial technology in smart production processes, offers an automated alternative and is increasingly applied in digital agriculture processes. Here we propose a method that combines YOLOv8 and morphological analysis of seeds using X-ray images to aid seed quality analysts in classifying seed lots based on their physiological potential. We examined the internal morphology of three seed lots and generated X-ray images to train a YOLOv8-based model with a post-segmentation module. This model assessed various parameters such as the total seed area, area filled by the embryo and endosperm, length, width, and calculated the percentage occupied by the embryo and endosperm in relation to the total seed area. Subsequently, seeds were classified into four categories based on the internal area occupied by the endosperm and embryo. Our findings reveal the robust performance of the YOLOv8 model in segmenting and classifying despite the challenges posed by a relatively small dataset. Notably, the proposed model achieved an impressive accuracy of up to 95.6% in identifying and segmenting the endosperm over 1,500 epochs with just 15 training images. The endosperm/seed area ratio, particularly within the 50-60% range covering over 50% of the samples, emerged as a significant metric for evaluating the viability of seed batches.

This article aims to present the development of a system of hardware and software to perform automatic data acquisition in photovoltaic panels for generating electricity. The system includes sensors, Hall Effect LA 55-P for the measurement of electric current and circuits for specific measures voltage, based on operational amplifiers. In addition to data acquisition for power, the system also performs the temperature monitoring panels, data obtained through these LM-35 sensor, and makes the presentation of information by means of graphs (generated in real time) and tables, both recorded in the database. The program is developed in C++ Builder, and tests have shown that the application appears as plausible tool for research on alternative energy, particularly in the PV, but also as an extremely powerful tool for monitoring the process of generating. O presente artigo tem como objetivo apresentar o desenvolvimento de um sistema de hardware e software para a realização automática de aquisição de dados em painéis fotovoltaicos para geração de energia elétrica. O sistema conta com sensores de Efeito Hall LA 55-P, para mensuração da corrente elétrica e circuitos específicos para medidas de tensão elétrica, baseados em amplificadores operacionais. Além da aquisição de dados para potência, osistema também realiza o monitoramento da temperatura dos painéis, dados estes conseguidos através do sensor LM-35, e realiza a apresentação das informações por meio de gráficos (gerados em tempo real) e tabelas, ambos gravados em banco de dados. O programe é desenvolvido em C++ Builder, e os testes realizados mostram que o aplicativo se mostra como ferramenta plausível para pesquisas na área de energias alternativas, em especial à fotovoltaica, mas também como uma ferramenta extremamente poderosa para supervisão do processo de geração.