With the appearance of Satellites and Drones, the remote sensing images are produced and have more attention in research. Image classification is one among its applications. The machine learning algorithms are widely used for this purpose in recent years. There are two main ways for classifying remote sensing images using machine learning techniques; the object-based image classification and the pixel-based image classification. This paper can be divided into two portions; the first portion presents a review of using machine learning in classifying the remote sensing images and the second portion transpires a comparative study using some carefully selected papers that surveyed in this field. The review illustrates the remote sensing images, the pixel-based image classification with some widely used algorithms, and the object-based image classification techniques, where the comparative illustrates a survey of some latest papers and make a comparative study between the support vector machine and the artificial neural networks algorithms which are utilized the object-based and the pixel-based image classification. This comparative study relies on measuring the overall accuracy and the kappa coefficient to assess the classifier performance. The experiment results demonstrate that the object-based image classification methods have higher performance than the pixel-based image classification methods for remote sensing image classification. In pixel-based image classification utilization, the experiment results also demonstrate that each of the support vector machine and the artificial neural networks algorithms performances can be more accurate than the other and vice versa according to research conditions.