Abstract

In the industry there is a strong requirement of automatic methods of quality control, and digital image processing meets this need very well. Therefore, the goal of this work is show the develoment of an IT tool for quality control in the production of eggs for consumption, which will lead to lower losses and increased's profit. The methodology adopted was structured and incremental development, where image processing techniques were used to classify eggs, on a treadmill, as good or bad. It was considered bad egg those that has dirt, stains, cracks or internal defects (blood spots). Is important observed that the developed tool does not distinguish between different types of defects because this would generate a waste of time in the classification. As a result it produces a hit rate of 86% in eggs with external defects and 66% in eggs with internal defects, overcoming various related works, and the time of classification is less that 3 seconds for each image.