INTELLIGENT OPTOMECHATRONIC INSTRUMENTATION FOR ONLINE INSPECTION OF CRUSHED ROCK AGGREGATES

Improved inspection of crushed rock aggregates will reduce the product ion of unsaleable material, enhance site quality control and boost sal es revenue through premium quality aggregate mixtures. A low cost opto mechatronic system is described which aims to provide a rapid breakdow n of particle size distribution, along with descriptions of constituen t product shape. A material sampling system directs aggregate into a n ovel optical inspection chamber, where particle geometry is digitised in real time by a dual beam laser scanning system. Classical image pro cessing algorithms obtain dimensional triaxial information from the di gitised data, and allow fractal or fourier features to be extracted if required. Soft computing techniques implement a neurofuzzy pattern cl assifier which assigns linguistic shape descriptors - rounded, irregul ar flaky etc. to the particle data. These methods combine to allow a d etailed and meaningful interpretation of the sampled crusher product.