Application of Colour Based Ore Sorting through Image Processing

摘要

Run- Off- Mine ore comprises of the valuable minerals and gangue materials which often vary in colours. To separate the ore from the waste and gangue materials Image Processing techniques are applied using live video of the ore and gangue mixture on the conveyorThe optical data captured by camera (sensor) is transferred via an optical data transferring system toa processor.The image processing techniques are applied to differentiate the ore and gangue material. The output signal is used to actuate appropriate r hanisms to direct the feed towards appropriate outgoing line. Thus the system reduces the,ore dilution and help in enhancing the ore grade. The video feed generated is converted to Hue Saturation Value (HSV) form id appropriate threshold is applied to differentiate ore from the impurities. Dynamic thresholding is done because this system is designed for onsite usage where the conditions will be changing and hence a flexible method is required. Detection by setting appropriate threshold in HSV format takes care of the ambient lighting conditions. The rrsults so obtained show the practicality and applicability of the method for ore sorting. This paper presents a laboratory scale development and testing of this technique and reports the output of the software developed in C for this purpose.