adopt two sensors and two switches to control the camera and
the gripper.
(2) The strawberry colour is a very important index to sort. The
Dominant Colour method fits to the human vision feeling, and
increase the grading precision, so that it is adopted in a* channel
to sort the strawberry colour feature. Based on this method,
three colour sorting indices were obtained.
(3) Because of the complex strawberry shape, the strawberry shape
classification algorithm was the main focus of this paper. The Kmeans
clustering method is the optimal algorithm in clustering
analysis. According to the advance class centers, this K-means
clustering method could put some similar class into one center.
So it is very fast for this method to calculate and classify.
In strawberry gradation, the strawberry shape is divided into
four classes, i.e., there are four advance class centers. The 14
lines were drawn from one edge of strawberry to the other,
and to be parameters of shape feature. The 14 lines were clustered
to classify the strawberry into four classes. The largest
fruit diameter was used to describe strawberry’s size.
(4) In order to carry out one, two or three indices to grade the
strawberry, the multi-attribute Decision Making Theory was
introduced. It can assign different weight to different attributes
of the object according to different standards to achieve the
simplification of the multi-attribute problems. According to the
grading require, the grading index and the weight of attribution
were entered to realize the multi-attribute strawberry classification.