Characterisation of flotation froth colour and structure by machine vision
Computers & Geosciences - Geological Applications of Digital Imaging
A Comparison of Algorithms for Connected Set Openings and Closings
IEEE Transactions on Pattern Analysis and Machine Intelligence
Hi-index | 0.00 |
Image denoising of coal flotation froth plays an important part in the subsequent image processing such as image segmentation and feature extraction. In traditional image denoising, there exists some inconsistency between removing the noise and preserving the most sharp detail information of object edges. In this paper, a morphological denoising algorithm is proposed for removing the noise of coal flotation froth image. This algorithm combines the opening and closing filters based on area reconstruction with an alternating order filtering method, and the elliptical structuring elements with increasing radius are adopted in the morphological filters. Based on the algorithm, denoise processing of a lot of coal froth images acquired from coal flotation working site was carried out. Denoising results show that many isolated spots on the original bubble images have been obviously eliminated, and no edge blurring appears, instead, the useful detail information in image is preserved.