This chapter highlights time series image processing for accurate agriculture characterization through two Moroccan experiences. The first case aims at crop mapping. A new classification approach based on multiple classifiers combination (MCC) was developed and applied to multi-temporal enhanced vegetation index (EVI) bands. The whole process is performed in three stages: (1) Landsat data preparation and multi-temporal staked EVI image extraction, (2) MCC construction from six advanced and supervised classifiers, and (3) stacked EVI image classification using the build-up MCC. Some post-classification contextual rules were also added in order to optimize the crops classification and the final parcel shape. In the second case, a post-classification change detection process was implemented to detect changes in forest area. Many classification schemes with different vegetation and texture indices were investigated. The two experiences are cost-effective, reproducible, and transferable. Consequently, they can regularly be used to produce up-to-date land use maps.

No abstract

No abstract

Remote sensing-based crop mapping has continued to grow in economic importance over the last two decades. Given the ever-increasing rate of population growth and the implications of multiplying global food production, the necessity for timely, accurate, and reliable agricultural data is of the utmost importance. When it comes to ensuring high accuracy in crop maps, spectral similarities between crops represent serious limiting factors. Crops that display similar spectral responses are notorious for being nearly impossible to discriminate using classical multi-spectral imagery analysis. Chief among these crops are soft wheat, durum wheat, oats, and barley. In this paper, we propose a unique multi-input deep learning approach for cereal crop mapping, called “CerealNet”. Two time-series used as input, from the Sentinel-2 bands and NDVI (Normalized Difference Vegetation Index), were fed into separate branches of the LSTM-Conv1D (Long Short-Term Memory Convolutional Neural Networks) model to extract the temporal and spectral features necessary for the pixel-based crop mapping. The approach was evaluated using ground-truth data collected in the Gharb region (northwest of Morocco). We noted a categorical accuracy and an F1-score of 95% and 94%, respectively, with minimal confusion between the four cereal classes. CerealNet proved insensitive to sample size, as the least-represented crop, oats, had the highest F1-score. This model was compared with several state-of-the-art crop mapping classifiers and was found to outperform them. The modularity of CerealNet could possibly allow for injecting additional data such as Synthetic Aperture Radar (SAR) bands, especially when optical imagery is not available.

The overall goal of this study is to define an intelligent system for predicting citrus fruit yield before the harvest period. This system uses a machine learning algorithm trained on historical field data combined with spectral information extracted from satellite images. To this end, we used 5 years of historical data for a Moroccan orchard composed of 50 parcels. These data are related to climate, amount of water used for irrigation, fertilization products by dose, phytosanitary treatment dose, parcel size, and root-stock type on each parcel. Additionally, two very popular indices, the normalized difference vegetation index and normalized difference water index were extracted from Sentinel 2 and Landsat satellite images to improve prediction scores. We managed to build a total dataset composed of 250 rows, representing the 50 parcels over a period of 5 years labeled with the yield of each parcel. Several machine learning algorithms were tested with the necessary parameter optimization, while the orthonormal automatic pursuit algorithm gave good prediction scores of 0.2489 (MAE: Mean Absolute Error) and 0.0843 (MSE: Mean Squared Error). Finally, the approach followed in this study shows excellent potential for fruit yield prediction. In fact, the test was performed on a citrus orchard, but the same approach can be used on other tree crops to achieve the same goal.