Abstract
Although there is much more data in RBG images, there is not often much more information than in a single band image. Where extraction of information is important, as opposed to visual fidelity, it is important to look first at the image capture stage, as enhancement here has many benefits. A system optimized by the SERD calculations described will contain more information (and less data) than the three channel RBG image, and so where colour differences are small, segmentation and object classification starting with such an image will be more accurate. Any established segmentation method can be used on the image captured with the optimized system, for example algorithms based on morphology or texture, and the accuracy and efficiency will be improved by the increase of relevant information
Published Version
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call