An apparatus for making a beverage including an infusion chamber for a capsule containing a food substance, an insertion opening for inserting the capsule into the apparatus, a transfer channel for transferring the capsule from the insertion opening to the infusion chamber, an image acquisition device to acquire at least one image of a portion of the capsule. The image acquisition device includes an optical sensor facing an image capture zone in which, in use, said capsule is located or passes. A viewing window, made of transparent material, is interposed between the optical sensor and the image capture zone. A heating element for heating the viewing window is applied to, or incorporated in, the viewing window.