Image processing based machine vision system for tomato volume estimation

Fleshy fruits are an essential part of the human diet providing vital vitamins, minerals and other health-promoting compounds. The quality of ripe fruit (such as texture, color, shelf life, sugar content) has significant effect on influences consumer acceptance, resistance against pathogens and transportability to long distances. Tomato is one of most largely grown agricultural products in the world, therefore, it is essential to deliver the fruit in high quality manner. There is a need of determining the quality attributes of this fruit (non-destructively) by fast and cost-effective techniques without damage due to the increasing demand of the in agro- industrially controlled areas. Most of the commonly employed techniques are time consuming and involve a considerable degree of manual work. Sample collection, cutting, grading and laboratory tests are among the limitations. Visual appearance is the main source of information about quality. This work aims to estimate volume of tomato, variety grown in Turkey, by image processing techniques. Five different images of a tomato are captured using high resolution digital cameras. Volume of the fruit is computed by estimating horizontal and vertical distance of captured images. The results are validated with experimental results. The main purpose of this study is to make fast and cheap determination of the fruit quality evaluation process without damaging the fruit and making it ready for packaging.