This paper describes an Infrared (IR) thermography based technique for sorting of iron ores consisting of alumina-rich particles of relatively low thermal absorptivity as compared to iron-rich particles in the ores. The technique primarily consists of selection of iron ores with Fe compositions ranging from 59 to 69 wt.% and alumina (Al2O3) from 0.35 to 8.85 wt.%, crushing the ores up to the particle size range around 10 mm. The iron ore fines are uniformly heated using heat source of wavelength ranging from 10(-2) to 10(-6) m for a period of time sufficient to create a difference in infrared emission between the ore particles. The thermal image of the heated ores is captured by IR thermography. The alumina-rich iron ore particles are heated up less as the thermal absorptivity of these ores is less than the ores with high iron content. Thus, the alumina-rich iron ore particles can be identified by observing the temperature profile and/or thermal image of these ores. This technique of ore recognition can be useful in improving the feed quality of iron ore to the blast furnace in iron and steel industries by rejecting the alumina-rich ores through modification in the presently existing processes. (c) 2013 Elsevier Ltd. All rights reserved.