Comparison of Speeded-Up Robust Feature (SURF) and Oriented FAST and Rotated BRIEF (ORB) Methods in Identifying Museum Objects Using Low Light Intensity Images

Abstract

Abstract Museum is a place of education and learning in the field of culture and history for all levels of society. As one of the first and largest museums in Lampung, Museum Lampung presents a variety of collections that are conditional on cultural values and are very useful if they can be identified through digital media. Speeded-Up Robust Feature (SURF) and Oriented FAST and Rotated BRIEF (ORB) methods are two examples of feature extraction methods that are relatively robust for object recognition in images by finding key points. Digital media determined by the value of key points in this study are images that are classified as having low intensity with an intensity value of <2250 Lux. This study compares the two methods using the digital object media of museum objects. The image of the museum object is given treatment in terms of rotation, scaling, and cropping to test the durability of the image matching process. In terms of feature matching time, the best time is achieved by SURF with 0.16 seconds in testing of 1/3 image region. Meanwhile, the highest matching percentage was also obtained by SURF method from rotational distortion at an angle of 90 degrees which is 76.79% instead of to 63.79% of the percentage obtained by ORB.