Potato is one of the world’s major crops and it is consumed daily by millions of people from diverse cultural backgrounds. Texture and colour are usually analyzed by conventional techniques and sensory panels which are expensive and time consuming. An interesting alternative to analyze colour and texture in food is to use computerized image analysis. The aim of the present research was to 1) Evaluate changes in texture and colour of processed potato applying image and instrumental techniques and 2) to establish relationship between images and instrumental analysis. Potato tuber cv. Spunta were cooked and freeze-dried. Microstructure (Scanning electron Microscopy), texture (Instrumental analysis (TPA) and image analysis) and colour (Spectrocolourimeter and image analysis) were analyzed. Similar structure was observed of cooked and freeze dried rehydrated samples. Statistical difference (P<0.05) was observed for texture and colour. In order to evaluate relation between image and instrumental analysis, data of texture and colour were subjected to linear normalization prior to PC analysis. A linear combination of selected principal components (PCs) was used to identity relationship between instrumental and image analysis. Each PC score was multiplied by the respective variance. For instrumental and image texture two PC´s was found accounting 90.7% and 81.7% of the total variance, respectively and two PC´s for colour accounting 96.7% (instrumental) and 95.1% (image). PC´s of texture and colour were combined using a linear combination, the linear regression analysis revealed a strong relationship between instrumental and image texture (R= 0.918) and for colour (R= 0.907), based on the resulting R2 value. The model showed that 84.0% (texture) and 87.0% (colour) of the variability associated image with instrumental analysis. Image analysis showed to be a useful tool due that it’s a non-destructive, economic and rapid method to estimate texture and colour.