A5080234713- Image Retrieval and Classification Techniques
- Advanced Image and Video Retrieval Techniques
- Medical Image Segmentation Techniques
- Business and Management Studies
- Industrial Vision Systems and Defect Detection
- Image and Object Detection Techniques
- Rough Sets and Fuzzy Logic
- Spectroscopy and Chemometric Analyses
- Postharvest Quality and Shelf Life Management
- Smart Agriculture and AI
- Data Management and Algorithms
- Digital Image Processing Techniques
- Advanced Database Systems and Queries
- Sensory Analysis and Statistical Methods
- Cancer-related molecular mechanisms research
- Plant and Fungal Interactions Research
- Solid State Laser Technologies
- Remote-Sensing Image Classification
- Meat and Animal Product Quality
- Food Drying and Modeling
- Genomics and Phylogenetic Studies
- Handwritten Text Recognition Techniques
- Collagen: Extraction and Characterization
- Machine Fault Diagnosis Techniques
- Image Enhancement Techniques
Universidade Nove de Julho
2015-2024
Universidade de São Paulo
2010-2017
Centro Latino-Americano e do Caribe de Informação em Ciências da Saúde
2008
This article presents a novel approach for detecting broken rotor bars in squirrel-cage induction motors, using time-domain current analysis. More particularly, this solution proposes new use of histogram oriented gradients since method is usually applied computer vision and image processing applications. Fully have been detected when the motor was running at very low slip operational condition difficult to identify traditional motor-current signature In addition, only one phase stator...
Machine learning algorithms have been applied to numerous transcript datasets identify Long non-coding RNAs (lncRNAs). Nevertheless, before these are RNA data, features must be extracted from the original sequences. As many of can redundant or irrelevant, predictive performance improved by performing feature selection. However, most current approaches usually select independently, ignoring possible relations. In this paper, we propose a new model, which identifies best subsets, removing...
In this work, we present the spectroscopic properties of LiYF4 (YLF) single crystals activated by high doping erbium ions. The most important processes that lead to up-conversion emissions in infrared region were identified. A time-resolved luminescence spectroscopy technique was employed measure decays and determine mechanisms involved populate S43/2 excited state. study energy transfer (ETU) Er:YLF showed an ETU rate can be obtained from I411/2 (ETU1) (ETU2) transient analysis, i.e., best...
In this study, seven wine samples were prepared varying the amount of pulp acerola fruits and sugar content using simulated annealing technique to obtain optimal sensory qualities cost for produced. S. cerevisiae yeast was used in fermentation process attributes evaluated a hedonic scale. Acerola wines classified as sweet, with 11°GL alcohol concentration aroma, taste, color characteristics fruit. The experiments showed that best conditions found at mass ratio between 1/7.5-1/6 total soluble...
In this work we propose a method for localizing text regions within scene images consisting of two major stages. the first stage, set potential is extracted from input image using residual operators (such as ultimate attribute opening and closing). second stage features obtained each region feature will be later used an to decision tree classifier in order label these or non-text regions. Experiments performed ICDAR public dataset show that good alternative problems involving location images.
This work introduces a residual operator called ultimate grain filter which is powerful image based on numerical residues. With multi-scale approach, the analyzes an under series of filters increasing sizes. Thus, contrasted objects can be detected if relevant residue generated when they are filtered out by one these filters. We also present efficient algorithm for computation using structure tree shapes. In fact, since result given obtained pruning corresponding and reconstructing it, we...
In this paper, we propose a collection of pattern sets to compute bit-quad matchings in component tree structures grayscale images. The proposed patterns are designed be used an efficient algorithm which runs linear time by counting incrementally each matching per pixel. It is worth emphasizing that the computes any attribute derived from 2×2 bit-quads such as area, perimeter, Euler number, among others. Finally, applied for circular object segmentation using shape ultimate opening with...
Current essay forwards a biodegradation model of dye, used in the textile industry, based on neural network propped by bootstrap remodeling. Bootstrapped is set to generate estimates that are close results obtained an intrinsic experience which chemical process applied. Pseudomonas oleovorans was reactive Black 5. Results show brief comparison between information estimated proposed approach and experimental data, with coefficient correlation real predicted values for more than 0.99 rate. Dye...
This work introduces a residual operator called ultimate attribute leveling. We also present an efficient algorithm for leveling computation by using structure tree of shapes. Our computating is based on the fact (proved in this work) that levelings can be obtained pruning nodes from novel result, since so far it known component trees. Finally, we propose use with shape information to extract contrast priori knowledge application. Experimental results applied text location show potentiality...