Size is one of the main fruit quality parameters. The customer preference internationally is for products that are sorted into sizes (Studman, 2001); the consumer subconsciously feels that if someone went to the trouble of uniformly sizing the produce, it must be of high quality (Peleg, 1985). The Marketing Standards for Fresh Fruit and Vegetables state that size is determined by the maximum diameter of the equatorial section or by the unit weight, depending on the product. Most modern sizers at fruit packinghouses use either machine vision, electronic weighing or both systems combined in the same machine. With the machine vision systems the product is classified according to its dimensional size, colour and sometimes external defects. The latter issue has a limited success, depending on the fruit and type of blemish [*]. Focussing in size, it is convenient to remark that machine vision equipment sometimes perform sorting errors. Even though size was one of the first parameters estimated automatically, there are still unsolved problems that can cause wrong estimations, mainly due to an excess of fruit in the sorter input (Aleixos et al., 2002). If we have to size e.g. grapefruit, wherein misshapen fruits (“sheepnosed”) often are present, many machine vision sizers will not classify them correctly. The reason is that a priori, taking into account the typical shape of this product, the user will configure the program to classify them according to their major axis, which corresponds to the equatorial in the case of grapefruits, but when a ‘sheepnosed’ fruit passes, the major axis will be the polar instead of the equatorial, and thus the sorting will fail. A way of overcoming this drawback would be the development of algorithms capable of recognizing the stem-calyx axis of the grapefruits. Because of the foregoing considerations it was considered interesting to keep on studying alternative low-cost systems for the fruit and vegetable sizing operation. One of them is the Optical Ring, matter of this report. We can mention some devices, somehow predecessors of the optical ring, since they are based on the same principle: duration of the light blocking, double pairs of transducers according to size ranges (Iwamoto and Chuma, 1981) and array phototransistors (Fon et al., 1990; Chen et al., 1992).
Size is one of the main fruit quality parameters. The customer preference internationally is for products that are sorted into sizes (Studman, 2001); the consumer subconsciously feels that if someone went to the trouble of uniformly sizing the produce, it must be of high quality (Peleg, 1985). The Marketing Standards for Fresh Fruit and Vegetables state that size is determined by the maximum diameter of the equatorial section or by the unit weight, depending on the product. Most modern sizers at fruit packinghouses use either machine vision, electronic weighing or both systems combined in the same machine. With the machine vision systems the product is classified according to its dimensional size, colour and sometimes external defects. The latter issue has a limited success, depending on the fruit and type of blemish [*]. Focussing in size, it is convenient to remark that machine vision equipment sometimes perform sorting errors. Even though size was one of the first parameters estimated automatically, there are still unsolved problems that can cause wrong estimations, mainly due to an excess of fruit in the sorter input (Aleixos et al., 2002). If we have to size e.g. grapefruit, wherein misshapen fruits (“sheepnosed”) often are present, many machine vision sizers will not classify them correctly. The reason is that a priori, taking into account the typical shape of this product, the user will configure the program to classify them according to their major axis, which corresponds to the equatorial in the case of grapefruits, but when a ‘sheepnosed’ fruit passes, the major axis will be the polar instead of the equatorial, and thus the sorting will fail. A way of overcoming this drawback would be the development of algorithms capable of recognizing the stem-calyx axis of the grapefruits. Because of the foregoing considerations it was considered interesting to keep on studying alternative low-cost systems for the fruit and vegetable sizing operation. One of them is the Optical Ring, matter of this report. We can mention some devices, somehow predecessors of the optical ring, since they are based on the same principle: duration of the light blocking, double pairs of transducers according to size ranges (Iwamoto and Chuma, 1981) and array phototransistors (Fon et al., 1990; Chen et al., 1992). Read More
Related posts:
Shape determination of horticultural produce using two-dimensional computer vision A review
Welch’s Absolute Fruitfuls Fruit Strips
Moosa dried-fruit teas
Aplicabilidad de un sistema de sensor de anillo óptico para determinación en línea del tamaño de frutos alargados y redondeados
Axis Communications unveils new experience center in Texas
Fruit size determination by a new optical sensor