JPS59230682A - Method of discriminating fruit - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

DETAILED DESCRIPTION OF THE INVENTION The present invention relates to a fruit-harvesting robot hand equipped with a fruit-harvesting device (2) as already proposed by the present applicant, for example in Japanese Patent Application No. 57-140021. The present invention relates to a fruit identification method that can be suitably applied to fruit guidance, etc.

In the above-mentioned robot hand for fruit harvesting, it is necessary to search and detect the position t of the fruit to be harvested, and then move the entire hand to a position K where the hand can capture the fruit.

However, in general, the position of fruit on the fruit tree is random, and the shape of fruit is also not constant, so a simple sensor cannot distinguish between fruit and other objects such as branches and leaves. is almost impossible.

Therefore, methods for identifying fruits using pattern recognition through image processing have been considered, but when applied to equipment used outdoors such as fruit harvesting equipment, the surrounding brightness is constant. Due to not having a single color filter etc.? With the general image processing used, it is extremely difficult to remove the background sky color and reflected light.

On the other hand, the above problem? Color image processing has been considered as a means of erasing the image, but it is not practical because the amount of information to be processed increases, the processing time becomes longer, and the equipment becomes bulkier. occurs.

The present invention has been made in view of the above-mentioned circumstances, and its purpose is to provide a fruit identification method that can easily remove the background sky and objects and obtain information corresponding only to the fruit.
i' is to provide.

In order to achieve the above object, the fruit identification method according to the great invention binarizes the difference signal between the red color signal and the blue color signal of the monitor camera. 5- to a corresponding to the fruit from the signal
All features should be extracted only.

The above configuration has brought about the following excellent effects.

In other words, in general, many fruits, such as tangerines, apples, persimmons, and strawberries, fall within the orange to red range, so what is the difference between the red signal and the blue signal among the three primary color signals? Just by calculating, you can reliably remove the background sky and reflected light regardless of the surrounding brightness, and also remove green light? Surrounding leaves can be easily removed just by not using it.
o Therefore, it has become possible to obtain an image signal corresponding to only fruits in real time at the image scanning speed of a monitor camera by simply performing a simple process of binarizing the difference signal between the red signal and the blue signal. be.

Embodiment 2 of the method of the present invention will be described below based on the drawings.

As shown in FIG. 1, three primary color signals (Rls tGl,
Color monitor camera fi that can be output separately from IBI
+ color signal (R1, (Gl), from FBI arithmetic unit (2
) to calculate the difference signal (R-B) between the red signal (R1 and the blue signal IBI) and compare this difference signal (R-g) with the reference signal (ref). (In order to convert to Dl, m is formed.

Then, the co-valued signal (Dl
By temporarily storing the fruit in a k-frame memory, etc., and calculating the center of gravity based on the address information, the entire fruit is separated from the image taken by the camera 7(x) and its position information is calculated. Therefore, it is used for @guiding the robot hand.

Incidentally, the monitor camera ftl does not use a color camera, but uses a normal monochrome camera and red and blue optical filters to substantially capture the red color signal fR1 in KIQ.
It is also possible to obtain an image signal corresponding to the blue signal IBI, temporarily store it in one frame memory, and then calculate the difference tone.

[Brief explanation of the drawing]

The drawing is a block diagram showing an embodiment of the fruit identification method according to the invention. m...Monitor camera, (R)...Red signal, En...Blue signal, (R-B)...
...Difference a.

Claims (1)

[Claims] By converting the difference signal (R-B) between the red signal (3) from the monitor camera (1) and the blue signal IBI into a t-binary value, the fruit is extracted from the image signal taken by the monitor camera + ILK. A fruit identification method characterized by extracting only signals corresponding to.