Current assembling lines for the fridge exploit a robotized assessment device that depends on cameras. As a developing issue, fridge arrangement dependent on pictures from its front view is important for cooler modern mechanization. In any case, this remaining parts an extremely testing task in light of the fact that the cooler is frequently seen against thick mess under conditions. In this paper, we propose a computerized fridge picture arrangement strategy, in light of the new undertaking of the Convolution Neural Network (CNN). It decides the challenges in fridge picture order by streamlining the information driven instrument and advancing both arrangement and comparability imperative. As far as anyone is concerned, this is the first occasion when that profound learning design has been reasonable for the fridge's home apparatuses field. Due to the investigations ordinary out utilizing 31,247 pictures of 30 classifications of coolers, our CNN engineering make a very monumental exactness of 99.96 %.