Currently, the word enclosed with Exabyte’s of digital information. The explosive growth of this content urged the need for developing a technique that is robust and efficient for detection the objects. In addition to detection, recognizing it despite of toughest conditions and image transformation in real time images. A lot of literature reviews and research works mention that in their works, the object detection is very significant process and used in many computer vision applications. Every work focused on detection the object under certain circumstances, and by using a methodology either concatenated with other technique or modified some approach, standard metrics or equations to achieve the desired goal. SURF algorithm used in many research works in object detection, and in most of these works, the algorithm achieves high accuracy. BRISK algorithm also used in wide applications, for object detection and recognition. Every algorithm has its own discriminative power to detect their interest points for forthcoming process. This paper focus on this approach, where the points that detected in both algorithms are used in proposed system instead single one. 27 Flicker logos dataset used to explore the accuracy of proposed system, and the results evaluated using four metric used commonly for evaluation in research works. Precision, Recall, Fmeasure and Accuracy have been shown the system achieve higher accuracy rate when compared with traditional systems and detect the objects even though the images are degraded significantly. However, the proposed system shown the combination of SURF and BRISK algorithms provide good results in detecting and recognizing the objects in degraded images.
Volume 10 | 02-Special Issue
Pages: 2330-2344