Prelimary Evaluation of a Computer Vision Sensor for Analysis of Phosphate Tailings

Computer control of the minerals beneficiation process is limited by a lack of adequate sensors to provide information on the current state of the process. Recent reduction in the cost of image acquisition and powerful inexpensive computers make use of computer vision as sensory input in process control a possibility. In the phosphate industry, the amount of apatite in phosphate tailings is estimated by inspection. For machine vision, a simple 256 x 512 color image of the tailings was captured via an inexpensive color video camera on an IBM PC. Initially, the image is converted to gray scale and displayed on an EGA color monitor. A program was written which identifies the different constituents of the sample by analyzing the gray scale levels for the digitized picture. In other instances the minerals are identified by their relative color intensities. The amount of each constituent is determined with good accuracy from the area it occupies on the image.