Machine Vision

Quality standards of fruit for fresh consumption used by electronic inspection systems are mainly based on the external appearance and absence of bruises, damages, and decay. The most used parameters to determine commercial categories are colour, size, shape, and visible damage inspected through high-speed machine vision systems. Images captured by these electronic devices must be processed in a few milliseconds to make decisions and sort the fruit in real-time. Additional challenges are related to the illumination system, which must be homogeneous, stable and uniform. The fruit is transported at a very high speed, which sometimes causes individualisation errors leading to misclassification. Moreover, the fruit is forced to rotate in the inspection area to capture views of most of the surface. Most of these machine vision systems are limited to the visible and near-infrared (NIR) regions of the electromagnetic spectrum, as NIR helps detect some invisible damage. However, other regions of the spectrum, such as ultraviolet (UV), may improve the accuracy of the detection of decay that is invisible to our eyes in the initial stages. This chapter summarises how electronic sorters inspect the fruit using computer vision systems, capturing images of the fruit in movement and extracting the most important external quality parameters using advanced image processing algorithms.