Automatic Assessment of River Water Level Images Using Machine Learning

الملخص

   River-level estimation is a critical task required for understanding flood events, but it is often complicated by the scarcity of available data. Recent studies propose using large networks of river-camera images to estimate river levels. However, this approach currently requires significant manual intervention, including ground topographic surveys and water image annotation. In this research, we present an innovative method to ease river-level estimation from river-camera images using through machine learning algorithms. In this project, the data cleaning process is done to remove any missing or distorted features or other anomalies in the data that need to be dealt with and is considered an initial stage of data processing. Then the stage of classifying the images into 2 categories. Based on the data set of the Kerala River in India, which is a unique set that includes a set of images taken for 2018 for all months. These images were processed and converted into digital data. This dataset contains 118 rows and 16 columns, including two columns named ANNUAL and RAINFALL. The RAINFALL column is the last completed column. Data can be classified as "yes" or "no" to determine whether the images in the dataset have been processed or not. Using K-Nearest Neighbor (KNN) algorithms, Random Forest (RF) and Support vector machine (SVM). Which achieved the highest percentage of 94.7%.