PENINGKATAN INTERPRETABILITAS CITRA DENGAN FUZZY LOGIC FUNGSI `S' UNTUK KEPERLUAN KLASIFIKASI

• Main Author: Ramdhan, Muhammad
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/5439
• Institution: Institut Teknologi Bandung

<p>Abstract:<p align=\"justify\"> <br /> Mapping technology by using remote sensing becoming the state of the art for latest period. It expected will be able to be done by fully automatic system, but there are still constraints in information extraction from image automatically, that is the subjectifity in interpretation. Image interpretability can be define as degree of coherence of object which still possible can be recognized by an image. The more sharply visual appearance of an image the more easy to interpretation and more information can be gain. To decrease the constrains there are many researchs in digital image processing that try to cope with it. In digital image processing we know that image enhancement is a process to improve the degree of interpretability, this is an alternative way to decrease the subjectifity. The process conducted by transforming appearance of histogram to obtain more contrast/sharp image such as in histogram equalization and linier stretching method. The idea is to change the distribution of brightness value with mathematical functions (linear, polinomial, etc). This research is about the implementation of fuzzy logic with \'S\' function to do the streching for a satellite image\'s histogram and compare its result with those conventional method in the domain of image classifications. The result shows that fuzzy logic with `S\' function can increase the KIA number for 12.37 % better than histogram equalization method and 31,13 % better than linier stretching method.<p align=\"justify\">