Several maize derived products such as tortillas (fried or without frying) are highly consumed in several Latin-American countries. The economic, cultural and social importance of these products is evident and engineering studies are relevant to design tortillas with better nutritional and sensorial quality. Computer vision is a valuable tool which helps to improve the automatic evaluation of food quality. The objective of this research was to study the relation between the oil content of tortillas chips under different processing conditions (time and pre-cooking) and some textural, chromatic and geometrical properties obtained from tortilla images by using computer vision. Tortillas of 2 mm of thickness prepared using nixtamalized moist dough called "masa" from maize were pre-cooked at 215°C during 1.5 min (pre-cooked tortillas). Tortillas without pre-cooking (control) and pre-cooked were fried at 175°C until reached a water content of 2% (wet basis). Digital images captured at each frying time were processed by computer vision. Surface oil was removed by immersion in petroleum ether for 10 seconds. Chips where superficial oil was removed were used to quantify structural oil content. Surface, structural and total oil content was determined using Soxhlet method and the results showed a lineal correlation (R > 0,95) between oil content (surface, structural and total) and some parameters of texture obtained by computer vision such as: Fourier, Hu and Haralick moments. A high correlation among different parameters of texture during time frying with oil content and oil distribution was observed In tortillas chips indicating that computer vision could be used to improve the quality evaluation of the frying process of tortilla chips.