Avaliação do índice de área foliar e índice de área da planta em floresta seca utilizando modelos simplificados em imagens de alta resolução com o uso de VANT
Bibliographic record
Abstract
O sensoriamento remoto tem possibilitado a aplicação de modelos para estimar variáveis ambientais, dentre eles o índice de área foliar (LAI) e o índice de área da planta (PAI), importantes para avaliação da sazonalidade da vegetação, principalmente em florestas secas. Assim, objetivou-se avaliar o LAI e PAI na caatinga usando imagens aéreas de alta resolução obtidas com um veículo aéreo não tripulado (VANT). Em área de caatinga preservada foram realizados voos com o VANT acoplado com câmeras RGB e RGN. Utilizou-se modelos para estimativa do LAI e PAI tendo como parâmetro de entrada o NDVI. Dados de LAI e PAI a partir do satélite Landsat-8 foram usados para comparação entre os produtos obtidos pelo VANT. A avaliação do NDVI ocorreu por regressão linear (R2=0,993), obtendo NDVI médio da Caatinga de 0,14 e 0,38 com os dados Landsat-8 nos períodos seco e chuvoso; 0,12 e 0,07 com a câmera RGB e RGN do VANT nos períodos seco e 0,65 e 0,27 para período chuvoso. Os dados LAI e o PAI (m2 m-2) representaram bem a área em estudo, obtendo R2=0,992 e R2=0,993 para LAI e PAI, respectivamente. O LAI médio da Caatinga foi 0,19 (período seco) e 0,80 (período chuvoso) pelo Landsat-8; 0,26 e 0,14 com a câmera RGB e RGN do VANT nos períodos secos e 2,18 e 0,48 para o período chuvoso. Já o PAI, os valores médios foram 1,39 e 2,02 com os dados Landsat-8 nos períodos seco e chuvoso; 1,46 e 1,34 com a câmera RGB e RGN do VANT nos períodos seco e 3,42 e 1,69 para o período chuvoso. Desse modo, os modelos calculados com imagens VANT para estimativa do LAI e do PAI da caatinga podem ser aplicados em imagens de alta resolução espacial obtidas em câmeras multiespectrais acopladas em VANT, obtendo resultados satisfatórios. Evaluation of leaf area index and plant area index in dry forest using simplified models in high resolution images using UAVRemote sensing has enabled the application of models to estimate environmental variables, including the leaf area index (LAI) and the plant area index (PAI), which are important for evaluating the seasonality of vegetation, especially in dry forests. Thus, the objective was to evaluate the LAI and PAI in the caatinga using high resolution aerial images obtained with an unmanned aerial vehicle (UAV). In a preserved caatinga area, flights were carried out with the UAV coupled with RGB and RGN cameras. Models were used to estimate the LAI and PAI having the NDVI as input parameter. LAI and PAI data from the Landsat-8 satellite were used to compare the products obtained by the UAV. The evaluation of the NDVI was carried out by linear regression (R2=0.993), obtaining an average NDVI of the Caatinga of 0.14 and 0.38 with the Landsat-8 data in the dry and rainy periods; 0.12 and 0.07 with the UAV's RGB and RGN camera in the dry season and 0.65 and 0.27 for the rainy season. The LAI and PAI data (m2 m-2) represented the study area well, obtaining R2=0.992 and R2=0.993 for LAI and PAI, respectively. The average LAI of the Caatinga was 0.19 (dry season) and 0.80 (rainy season) by Landsat-8; 0.26 and 0.14 with the UAV's RGB and RGN camera in the dry season and 2.18 and 0.48 for the rainy season. As for the PAI, the average values were 1.39 and 2.02 with the Landsat-8 data in the dry and rainy seasons; 1.46 and 1.34 with the UAV's RGB and RGN camera in the dry season and 3.42 and 1.69 for the rainy season. Thus, the models calculated with UAV images to estimate the LAI and PAI of the caatinga can be applied to high spatial resolution images obtained from multispectral cameras coupled to UAVs, obtaining satisfactory results.
Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.
How this classification was reachedexpand
Full frame machine prediction
Teacher imitationNot calibrated prevalence, not ground truth. Human validation pending. The Gemma side is a direct model label for every work in the frame, read from the title-only record. The Codex side is a classifier learned from the 10,348 direct Codex labels and calibrated to design-weighted sample rates; fields without enough sample support carry no Codex call. Candidate is the union of the two sides; consensus is their intersection. These outputs are machine_predicted_unvalidated and are not human labels.
Distilled classifier scores by category (both heads)
| Category | Codex | Gemma |
|---|---|---|
| Metaresearch | 0.000 | 0.001 |
| Meta-epidemiology (narrow) | 0.001 | 0.000 |
| Meta-epidemiology (broad) | 0.001 | 0.001 |
| Bibliometrics | 0.001 | 0.001 |
| Science and technology studies | 0.000 | 0.000 |
| Scholarly communication | 0.001 | 0.001 |
| Open science | 0.000 | 0.000 |
| Research integrity | 0.001 | 0.000 |
| Insufficient payload (model declined to judge) | 0.003 | 0.001 |
Machine scores (provisional)
The two teacher heads of the student model, read on this work. A score orders the frame for review; it never asserts a category, and the validation status ships verbatim with every row.
Baseline scores from an immature model (maturity gate not passed, 7 training rounds). Scores rank; they never assert a category.
score_only:v0-immature-baseline · verbatim from the scoring run: score_only means the number may rank works, and no category label ships from itClassification
machine, unvalidatedMachine predicted; a candidate call from one source (direct Gemma or distilled Codex), not a consensus.
How this classification was reached, model by model and score by score, is at the end of the page under "How this classification was reached".