Evaluating a Machine Vision System by Measuring Some Physical Characteristics of Pistachio Nuts

In order to increase the role of machine vision in agricultural research in Iran, especially for measuring physical attributes of seeds, a machine vision system was developed using a computer, a capture card, a video camera and a light box. All equipment was purchased from domestic markets. Computer programs were developed for hardware setup and for image processing applications. The programs perfomed tasks such as image acquisition and display, color conversion, image segmentation, object counting, and measurement of some physical attributes of the objects by analysing their images. The system was used to measure some physical attributes of pistachio nuts. The machine vision measurements were statistically compared with the measurements obtained by the conventional manual methods. The results indicated that there was generally no significant difference between the two methods. However, the time consumed by the machine vision method was far less than the time taken by manual methods. The experimental results also showed that there were many sources of error and limiting factors in using machine vision for measuring physical attributes of seeds.