Abstract

Water is one of the unique objects in nature. In computer vision applications, object identification mechanism of water can be done by analyzing the light on the object, either by analyzing the reflection or absorption of light. In addition, the approach in the detection of surface water must also consider the adequacy of the light source and some properties of water associated with the propagation of light, ie the water surface conditions (calm, wavy), the status of depth, turbidity, and presence or absence of material that floats on the water surface such as foam or other materials). The nature of this water will affect the reflection or absorption of light, which would affect the number of variations that must be defined for the detection of water. Several research approaches for detecting the water level that has been done is a color image classification method with color imagery and analysis of light polarization. This research is a study of research literature related to the mechanism of water detection using 2 methods. Discussion is to do the comparison includes comparing the descriptions of equipment and removal of surface water scenario based approach to the properties of water. Then, the comparison of image segmentation is arranged based on the said pattern identification.