JPH0268184A - Sorter for fruits - Google Patents

Abstract

PURPOSE:To perform automatic assortment by judging an imaging state as an abnormal one wherein the images of two or more fruits or vegetables are taken when the shape taken by a shape imaging part has a length equal to or more than a predetermined reference length and performing estimation on the basis of a predetermined standard to classify fruits or vegetables. CONSTITUTION:A classification judge part 9 reads the binarization signal of the object to be sorted from a shape imaging part 2 in predetermined timing through a data collection part 8. Whereupon, predetermined processing such as the removal of a noise signal or the extraction of characteristic is performed with respect to the binarization signal to be stored temporarily. Subsequently, the classification judge part 9 uses the binarization signal stored temporarily to calculate not only the imaging lengths H, W in the horizontal and vertical directions of the object to be sorted but also oblateness R and compares said oblateness R with set oblateness Rmax and judges that the object to be sorted is not a nonstandard one at the time of R<=Rmax to advance to the next step of sorting processing. By this method, efficient assortment is performed.

Description

DETAILED DESCRIPTION OF THE INVENTION [Field of Industrial Application] The present invention relates to a sorting device for sorting fruits and vegetables according to set criteria.

[Prior Art] With the development of pattern recognition technology in recent years, inventions in various fields of this technology are flourishing. For example, when it comes to sorting fruits and vegetables such as mandarin oranges and apples, automatic sorting using pattern recognition has begun to be used to correct variations in sorting standards and improve efficiency caused by conventional manual sorting. ing.

[Problems to be Solved by the Invention] Incidentally, in such automatic sorting, it is fundamental that each object to be sorted can be reliably imaged and recognized individually. In other words, if a plurality of objects to be sorted enter the imaging area in such a way that they are overlapped and are imaged, performing recognition processing on these objects as - objects to be sorted will naturally result in incorrect sorting results. This is because there is no In particular, in the automatic sorting of fruits, vegetables, etc. mentioned above, when the objects to be sorted are approximately spherical, such as oranges and apples, the objects to be sorted that constitute the automatic sorting line are transported to the imaging area. On the way, the supplies to the conveyance path may be inaccurate, or they may roll during conveyance, resulting in overlapping conveyance to the imaging area.

Japanese Unexamined Patent Application Publication No. 1983-1989 is a conventional technology that solves these problems.
There is a publication No. 266188. In this conventional technology, when a plurality of objects to be sorted are overlapped and imaged, the plurality of objects to be sorted are set in an abusive imaging state, are rendered unsortable, fall from the conveyance path, and are transported again by the imaging device. It is returned to the upper half, imaged again, and sorted. Therefore, although the accuracy of sorting the objects to be sorted is improved, there is a problem in that the sorting efficiency is reduced.

Therefore, the present invention aims to solve these problems.

[Means for Solving the Problems] The technical means of the present invention for solving such problems is to place fruits and vegetables on a conveyor bag GLI at a distance from each other in the front and back, This is a sorting device that images the shapes of fruits and vegetables with a shape imaging unit 2 during the conveyance process, and discharges and sorts fruits and vegetables according to predetermined shapes. If the above is the case, it is determined that there is an abnormal imaging state in which two or more fruits/vegetables are imaged, and then the individual shapes of these two or more fruits/vegetables are estimated based on a predetermined standard, and the sorting and discharge are determined. The present invention is characterized by the configuration of a sorting device for fruits, etc.

[Operations and Effects] Even in an abnormal imaging state where multiple objects to be sorted are imaged overlappingly, the shapes of the multiple objects to be sorted are estimated, the shapes are discriminated, and the classification is performed, so the discrimination accuracy slightly decreases. However, it is possible to carry out efficient sorting work by omitting the need to transport the objects to be sorted back to the conveyance start end side and take an image of the shape again.

[Example] Embodiments of the invention shown in the drawings will be described below.

FIG. 1 shows a sorting device showing one embodiment of the present invention, which is particularly suitable for sorting roughly spherical fruits such as oranges and apples, or potatoes with a generally circular cross section and long lateral sides. be. Reference numeral 1 denotes a so-called keyboard-type conveyance device that conveys the objects to be sorted at a predetermined speed, 2 a shape imaging section that images the shape of the objects to be sorted that are conveyed by the conveyance device 1, and 3 a shape imaging unit that images the shape of the objects to be sorted that are conveyed by the conveyance device 1. This is a discharge section that is provided on the downstream side of the section 2 in the conveyance direction and discharges the conveyed objects to be sorted from the conveyance device 1 in response to a drive signal from a signal processing section, which will be described later.

The conveying device 41 has a plurality of regions (hereinafter referred to as "divided regions") divided in the width direction by convex portions provided at intervals of a certain distance d corresponding to the length of the imaging region in the shape imaging section 2. The size of each divided area is large enough to accommodate one object to be sorted, regardless of its size.

Therefore, in some cases, due to a malfunction of the supply device (not shown) that supplies the materials to be sorted at the starting end side of the conveyance device 1, one
Instead of each object being sorted a predetermined distance apart, subsequent objects to be sorted may be stacked one on top of the other when viewed from the left and right sides, and the objects to be sorted may be placed in one section. A situation may occur in which a plurality of devices are placed in a region and images are taken at the same time.

The shape imaging unit 2 has a camera formed of, for example, an area image sensor, which is provided at positions facing each other across the transport device 1 in the width direction thereof, and a light source formed of, for example, a fluorescent lamp. With respect to the objects to be sorted that have been conveyed on the conveyance device 1, the shape of the objects to be sorted is imaged for each division area using a so-called transmission method, and the binarized signal is outputted to the signal processing section 4. It is something.

The signal processing section 4 performs sorting judgment on the objects to be sorted imaged from the binarized signal sent from the shape imaging section 2 based on preset sorting criteria, and according to the result, performs sorting judgment for each sorting criterion. Gate section 3- constituting the discharge section 3 of the sorted material
It outputs a drive signal to 1 to 3-n.

The specific configuration includes an encoder section 6 that outputs a pulse signal according to the rotational speed of a rotating body 5 provided in contact with the conveyance device 1 (conveyance speed of the conveyance device 1), and a Upon receiving the pulse signal, the shape imaging section 2. A synchronous controller unit 7 outputs necessary activation signals to a data collection unit 8 and a gate control unit 10, which will be described later, and a data collection unit appropriately collects and outputs binary signals from the shape imaging unit 2 according to the synchronization signals. section 8, a classification determination section 9 that performs the aforementioned sorting and determination based on the binary signal supplied via this data collection section 8, and a classification determination section 9 that performs the above-mentioned sorting determination based on the binarized signal supplied via this data collection section 8; It has a gate control section 10 that outputs a drive signal at an appropriate timing according to the synchronization signal from the synchronization controller section 7. In addition, in the synchronous controller section 7, one
The classification determination unit outputs a synchronization signal for starting imaging each time one divided area is sequentially set in the imaging area, and at the same time stores the binarized signal of the previous object to be sorted to the data collection unit 8. 9, and outputs a synchronization signal for newly storing the binarized signals from the shape imaging section 2 and the like. Further, the synchronization controller section 7 outputs a synchronization signal indicating the timing for determining the current position of each sectioned area of the transport device 1 to the gate control section 10.

Next, the principle of detecting an abnormal imaging state in the classification determining section 9 will be explained using the second p2I. The abnormal imaging state is basically determined based on the magnitude of the flattening ratio R (=H/W) of the object to be sorted, which is imaged based on the binary signal. In other words, if the type of object to be sorted is clear, the flatness ratio R of this object is
Since the upper limit (Rmax) of (1) can be approximately estimated, for example, if an image within the normal standard is normally imaged as shown in FIG. 2(a), R) will not reach Rmax. However, if the object has an aspect ratio R that exceeds this upper limit (Rmax), it may be in an abnormal imaging state where multiple objects are overlapped, as shown in FIG. As shown in Figure 2 (c), it can be determined that the object to be sorted is out of the standard, and in particular, in the latter case, R>Rmax holds true no matter how many times the image is taken, so if this is excluded. The rest can be determined to be in an abnormal imaging state.

In the case of an abnormal imaging state, the shape of the object to be sorted is estimated and the object is estimated to meet a predetermined sorting criterion, and a sorting determination is made in the next step. That is, (2) the minimum width portion (h in FIG. 2) that is less than or equal to the predetermined length is measured at a predetermined length of the intermediate portion along the conveyance direction in the abnormal imaging state.

In addition, as shown in FIG. 2 (Q), if the minimum width less than the predetermined length is not measured at the predetermined length of the intermediate part along the conveyance direction, the entire object is classified into one piece. It is judged as non-standard.

■The part of the object to be sorted on the upper side of the conveyance side than the minimum width part is the outer part of the object to be sorted which has a curvature smaller by a predetermined percentage of the part of the part of the object to be sorted on the lower side of the conveyance side than the minimum width part that connects to the minimum width part. The point where the upper and lower contour lines intersect is estimated to be the tip of the lower side of the conveyed object, and the shape is estimated and determined. ,
The part on the upper side of the conveyor than the minimum width part is assumed to be the outline of the object to be sorted which has a curvature smaller by a predetermined percentage than the outline of the object to be sorted that connects to the minimum width part, and the point where both the upper and lower outlines intersect is It is estimated that this is the tip of the upper part of the conveyed object, and the shape is estimated and determined.

(2) They are classified based on the above-mentioned shape estimation and selection criteria, and drive signals are output to the gate sections 3-1 to 3-n constituting the discharge section 3.

The discharge section 3 has a plurality of gate sections 3-1 to 3-n for discharging the objects to be sorted from the conveyance device 1, and the number of gate sections 3-1 to 3-n are provided in a number corresponding to the sorting standards (including non-standard) of the objects to be sorted. The material to be sorted is discharged according to the sorting result in the classification determining section 9 and supplied to another transport route (not shown).

Next, the operation of this embodiment will be explained.

During the sorting work, the classification determination unit 9 reads the binary signal of the object to be sorted from the shape imaging unit 2 via the data collection unit 8 at a predetermined timing, and first detects a noise signal for this binary signal. Performs preprocessing such as removal of , feature extraction, etc., and stores the result at - time.

(2) Next, the first classification determining section 9 determines the horizontal and vertical imaging lengths H and W of the object to be sorted using the stored binary signal.

■Next, from this result, the flatness ratio R (=H
/W) is calculated.

■The classification determination unit 9 then compares the calculated flatness ratio R with the set flatness ratio Rmax, and if R≦Rmax holds, the imaged object to be sorted is in an abnormal imaging state as shown in FIG. 2(a). It is determined that the product is not outside the standard, and it is not outside the standard.
Proceed to the next step of sorting.

■On the other hand, if R>Rmax is established, the imaged object to be sorted is judged to be in an abnormal imaging state or non-standard, as shown in (b) or (Q) in Figure 2, and the next step is taken. Proceed to step.

Next, if there is no abnormal imaging state, the classification determination unit 9 further classifies the object to be sorted using the binarized signal according to nine criteria preset based on size, shape, etc. , performs inspection and selection, and outputs a command signal to the gate control unit 10 to drive one of the gate units 3-1 to 3-n corresponding to the selection result.

■Next, the classification determination unit 9 first determines that the object to be sorted has the same result as the previous step, that is, the flatness ratio R is set to the set flatness ratio R.
It is determined whether or not it is determined that the value is larger than max. Then, it is detected that the same result was obtained in the previous stage for this object to be sorted, and the minimum width portion less than the predetermined length is detected between the predetermined length of the intermediate portion along the im feed direction. If not, the classification determination unit 9 determines that the object to be sorted has an abnormally large flattening ratio R as shown in FIG. A command signal is output to the gate control section 10 to drive the corresponding gate section.

■Also, at the stage of ■, the flatness ratio R is the set flatness ratio Rm
ax, and if there is a minimum width gap that is less than or equal to the predetermined length between the predetermined length of the intermediate portion along the conveyance direction, the classification determination unit 9, for example, Figure 2 (
As shown in b), it was determined that the object to be sorted was in an abnormal imaging state.

a: The part of the object to be sorted on the upper side of the conveyance side than the minimum width area has an outer shape with a curvature that is smaller by a predetermined percentage of the curvature of the outline of the object to be sorted that connects the part of the object on the lower side of transport than the minimum width area to the minimum width area. The point where the upper and lower contour lines intersect is estimated to be the tip of the lower side of the conveyed object, and the shape is estimated and determined. , the part on the upper side of the conveyor than the minimum width part is assumed to be the part with a curvature smaller than the outline of the object to be sorted by a predetermined percentage that connects to the minimum width part, and both the upper and lower outlines are The intersection point is estimated to be the tip of the upper part of the object to be transported, and the shape is estimated and determined.

b=The objects to be sorted are classified according to the above-mentioned shape estimation and sorting criteria, and a drive signal is output to the gate section constituting the discharge section 3.

As a result of such processing in the classification determination unit 9, even in an abnormal imaging state where multiple objects to be sorted are imaged overlappingly, the shapes of the multiple objects to be sorted are estimated and discriminated, so the discrimination accuracy is improved. Although the efficiency is slightly lowered, it is possible to omit the return conveyance of the objects to be sorted to the conveyance start end side and perform efficient sorting.

It is a front view.

Explanation of symbols Conveyance device 2 Shape imaging unit Discharge section 3-1 to 3n...gate section Signal processing section 5 Rotating body Encoder section Synchronous controller section Data collection section 9 Classification judgment section Gate control section

Claims (1)

[Scope of Claims] [1] Fruits and vegetables are placed on the conveyance device 1 at a distance from each other in the front and back, and the shapes of the fruits and vegetables are imaged by the shape imaging section 2 during the conveyance process of the conveyance device 1. , a sorting device that sorts, discharges, and sorts each predetermined shape, and when the shape imaged by the shape imaging unit 2 has a predetermined reference length or more, an abnormal image in which two or more fruits/vegetables are imaged is detected. A sorting device for fruits, etc., characterized in that the fruit/vegetables are sorted and discharged by determining the state of the fruit/vegetable, and then estimating the shape of each of the two or more fruits/vegetables based on a predetermined standard.