Extraction of Arecanut Planting Distribution Based on the Feature Space Optimization of PlanetScope Imagery

The remote sensing extraction of large areas arecanut (Areca catechu L.) planting plays an important role in investigating the distribution area and subsequent adjustment optimization regional structures. Satellite imagery has previously been used to investigate monitor agricultural forestry vegetation Hainan. However, monitoring accuracy is affected by cloudy rainy climate this region, as well high level land fragmentation. In paper, we PlanetScope at a 3 m spatial resolution over Hainan high-precision based on feature space optimization. First, spectral textural variables were selected form initial space, followed implementation random forest algorithm optimize space. Arecanut models support vector machine (SVM), BP neural network (BPNN), (RF) classification algorithms then constructed. overall accuracies SVM, BPNN, RF optimized features determined 74.82%, 83.67%, 88.30%, with Kappa coefficients 0.680, 0.795, 0.853, respectively. model exhibited highest kappa coefficient. following was improved 3.90%, 7.77%, 7.45%, respectively, compared corresponding unoptimized model. coefficient also improved. results demonstrate ability satellite extract arecanut. Furthermore, proven effectively composed variables, further improving distribution. This work can act theoretical technical reference for industries.