| Summary: | 碩士 === 國立高雄應用科技大學 === 電子工程系 === 106 === In recent years, image stitching has been widely discussed in computer vision. In the applications of driver assistant and surveillance systems, image stitching is one of the most important components. Feature extraction and classification are two main steps in image stitching. ORB is an algorithm used in computer vision to detect and describe images. The algorithm consists of the following two steps: (1) Oriented FAST and (2) Rotated BRIEF. Next, we match and adapt the feature points through RANSAC (random sample consensus), then project the different images into the same flat by cylindrical projection. Finally, multi-band blending is used to deal with the overlap of images to eliminate the difference between images. Consider the parallel computing and heterogeneous computing of OpenCL on the various embedded platforms, we can improve and accelerate our image stitching method. Finally, an ORB-based Image Stitching Method by Open Computing Language for the embedded platform is proposed.
|
|---|
