Sequential Categorization of Date Palm White Scale Disease via Machine Learning Methods
摘要
In the regions of North Africa and the Middle East, the cultivation of date palms assumes a pivotal role not only in environmental preservation but also in driving economic development, particularly within predominantly oasis regions. The date palm, with its remarkable ability to combat desertification through the interception of intense solar radiation and the establishment of a veritable “green and productive wall,” stands as a vital guardian of these ecosystems. In recent times, the Artificial Intelligence (AI) market within the realm of precision agriculture has surged to prominence, representing a substantial component of the Software and Services economy. This ascent is underscored not only by the magnitude of investments but, more fundamentally, by the profound innovations and technological advancements it has ushered in. In the dynamic field of agricultural technology, we are committed to fully deploying modern AI instruments to advance the cultivation of date palms, with an emphasis on improving the detection of Date Palm White Scale Disease. We have developed an extensive framework anchored in the solid principles of machine learning and advanced methods for extracting features. This strategy enhances our ability to assess the extent of Date Palm White Scale Disease infestation. By integrating the strengths of Hu Moments, Haralick Textures, and Color Histograms for feature extraction, and employing the robust Random Forest Classifier for the classification process, we have attained an exceptional average accuracy rate of over 99%.