CN116801993A - Optical sorting machine - Google Patents

Abstract

The optical sorter includes: a light source; an optical sensor configured to detect light associated with a sort of material; a detection unit configured to detect a defective portion of the sorted object based on the signal acquired by the optical sensor; a sorting section configured to sort the sorted objects having the reject portion by injecting air toward the reject portion; and an injection control unit configured to control the injection of air. The injection control unit is configured to execute injection range control in which air is injected into the 1 st injection range with respect to the sorted object when the value of the size of the reject portion is the 1 st value, and air is injected into the 2 nd injection range with respect to the sorted object when a predetermined condition is satisfied. The 2 nd injection range is wider than the 1 st injection range. The prescribed condition includes that the value of the size is a2 nd value smaller than the 1 st value.

Description

Optical sorting machine

Technical Field

The present invention relates to an optical sorter.

Background

Conventionally, an optical sorting machine is known which uses optical information obtained by an optical sensor when light is irradiated from a light source to an object to be sorted to discriminate and remove foreign matters or defective products contained in the object to be sorted. The optical information (for example, a tone value) obtained by the optical sensor is compared with a threshold value, and whether the sorted object is a good or a foreign object or a bad is determined based on the comparison result. The sorted objects determined to be the foreign objects or the defective objects are typically blown off by air jets, whereby the sorted objects are sorted into the defective objects, the foreign objects, and the defective objects.

With respect to such an optical sorter, japanese patent No. 4206522 discloses a technique in which the ejection period of air is set longer as the width of the reject portion of the sorted object is larger. In this technique, the sorted objects having the reject portions with a large width are estimated to be large particles (in other words, particles with a large weight), and the particles that are difficult to blow due to the large weight are intended to be reliably blown off during the ejection period set longer.

However, the optical sorter described above has room for improvement in order to improve sorting accuracy.

For example, in the case where the sorted object is rice grains and there is a small colored portion at the end of the rice grains, in the above technique, the injection period of air is set to be short. In this case, if air is injected with respect to the colored portion, the injected air may contact the end of the rice grains in a small amount, and the trajectory of the rice grains may not be changed to a sufficient degree to be distinguished as a good product (hereinafter, also referred to as a small contact phenomenon). This phenomenon results in a decrease in sorting accuracy.

On the other hand, when the center of the whole rice grains having the colored portion is detected and air is injected with the center as a target, the above phenomenon is less likely to occur. However, since the center detection of rice grains requires complicated signal processing, the load of the signal processing increases, and as a result, the processing time may increase. Further, the center detection of the rice grains requires detection of the outline of the rice grains, but when a plurality of rice grains are transferred in a state of being in contact with each other, it is difficult to accurately detect the outline of 1 grain. If the outline of the rice grains cannot be detected accurately, the sorting accuracy is lowered.

Disclosure of Invention

The present disclosure has been made to solve at least a part of the above problems, and can be realized, for example, as follows.

According to mode 1 of the present disclosure, there is provided an optical sorter. The optical sorter includes: a light source configured to irradiate light to the conveyed objects to be sorted; an optical sensor configured to detect light irradiated from a light source and associated with a sort object; a detection unit configured to detect a defective portion of the sorted object based on the signal acquired by the optical sensor; a sorting unit configured to sort the sorted objects having the reject portion by injecting air toward the reject portion; and an injection control unit configured to control the injection of air. The injection control unit is configured to execute injection range control in which air is injected into the 1 st injection range with respect to the sorted object when the value of the size of the reject portion is the 1 st value, and air is injected into the 2 nd injection range with respect to the sorted object when a predetermined condition is satisfied. The 2 nd injection range is wider than the 1 st injection range. The prescribed condition includes that the value of the size is a2 nd value smaller than the 1 st value.

The "light associated with the object to be sorted" may be reflected light that is light reflected by the object to be sorted, transmitted light that is light transmitted through the object to be sorted, or both reflected light and transmitted light. The "reject portion" refers to one or both of a portion of poor quality (here, the whole of the sorted object or a part of the sorted object) and a foreign matter (in this case, the whole of the foreign matter is a reject portion). The "dimension" of the reject portion refers to any one of the area, length, and width of the reject portion or any combination thereof. The "injection range" of air with respect to the objects to be sorted is not limited to a range in which air is injected at a certain instant, and refers to a range in a coordinate system that moves together with the objects to be sorted while being transferred. For example, the relative positional relationship between the object to be sorted and the region where the air is injected during the transfer varies from time to time, but in a coordinate system in which the object to be sorted moves together, the whole of the coordinate region where the air is even momentarily in contact may become the "injection range" of the air with respect to the object to be sorted.

According to this optical sorter, when the predetermined condition is satisfied, the air injection range with respect to the sorted object is set wider when the size of the reject portion is a2 nd value greater than the 1 st value than when the size of the reject portion is the 1 st value. In other words, when the reject fraction is small to a predetermined extent, the injection range of air to the sorted object is widened. Therefore, in the case where a small reject portion exists at the end of the sorted object, even if air is ejected toward the reject portion, a small contact phenomenon is less likely to occur. Therefore, the sorting accuracy can be improved.

According to claim 2 of the present disclosure, in claim 1, the injection range control includes a1 st control in which, when the value of the dimension is the 2 nd value, the injection period of the air is set to a longer time than when the value of the dimension is the 1 st value. According to this aspect, the contact range between the air in the transfer direction of the objects to be sorted and the objects to be sorted can be widened, and therefore, the small contact phenomenon is less likely to occur. Therefore, the sorting accuracy can be improved. The 1 st control can be understood as control for changing the injection range by changing the injection period of air.

According to claim 3 of the present disclosure, in claim 1 or 2, the sorting section includes a plurality of nozzles that can selectively eject air and are aligned in a direction orthogonal to the transfer direction of the objects to be sorted, that is, in an orthogonal direction. The injection range control includes a2 nd control in which, when the value of the size is the 1 st value, air is injected from 1 nozzle among the plurality of nozzles, and when the value of the size is the 2 nd value and a predetermined portion (for example, an arbitrary portion or a center) in the reject portion is separated by a predetermined distance or more from the center of the injection range of air in the orthogonal direction of the 1 nozzle toward the orthogonal direction, air is injected from 1 nozzle among the plurality of nozzles and the nozzle adjacent to the 1 nozzle. According to this aspect, when the defective portion is small to a predetermined extent and the predetermined portion of the defective portion is present in the vicinity of the boundary of the injection range in the orthogonal direction of each of the adjacent 2 nozzles, air is injected from the adjacent 2 nozzles, and the contact range between the air in the orthogonal direction and the sorted object can be widened, so that the small contact phenomenon is less likely to occur. Therefore, the sorting accuracy can be improved. The 2 nd control can be understood as control for changing the injection range by changing the number of nozzles to which air is to be injected among the plurality of nozzles.

According to claim 4 of the present disclosure, in claim 3, each of the plurality of nozzles variably corresponds to an injection charge range in the orthogonal direction with respect to each detection position of the sorted object. The injection control unit is further configured to control the injection of air so that each of the plurality of nozzles injects air when a predetermined portion of the defective portions is within a corresponding injection charge range. The plurality of injection responsible ranges corresponding to the plurality of nozzles are set so as not to overlap each other when the value of the dimension is the 1 st value, and are set so that any adjacent 2 of the plurality of injection responsible ranges partially overlap each other when the value of the dimension is the 2 nd value. When a predetermined portion of the reject portion is located in the overlapping region of the adjacent 2 injection responsible regions, air is injected from 2 nozzles corresponding to the adjacent 2 injection responsible regions. According to this aspect, the 3 rd aspect can be realized by a simple operation.

According to claim 5 of the present disclosure, in any one of claims 1 to 4, the detecting unit is further configured to detect a center of the defective portion. The injection control unit is further configured to perform injection range control with reference to the center of the defective portion. According to this aspect, the injection range control can be performed with higher accuracy.

According to a 6 th aspect of the present disclosure, in a 5 th aspect including the 2 nd aspect, the 1 st control includes setting timings of starting and ending the injection period so that the timing of injecting air to the center of the defective portion is located at the center of the injection period. According to this aspect, when the reject fraction is located on one side in the transfer direction of the sorted objects, air can be reliably brought into contact with the sorted objects having the reject fraction regardless of which side in the transfer direction the reject fraction is biased. Therefore, the sorting accuracy can be improved.

According to a 7 th aspect of the present disclosure, in the 5 th aspect including the 3 rd aspect or the 6 th aspect including the 3 rd aspect, the predetermined portion of the reject portion is a center of the reject portion. According to this mode, opportunities for injecting air from 2 nozzles adjacent to each other are defined. Therefore, the air ejected from the 2 nozzles can be suppressed from changing the trajectory of other sorted objects in the vicinity thereof together with the sorted objects having the reject portions. Thus, the yield is improved.

According to an 8 th aspect of the present disclosure, in the 3 rd or 4 th aspect, the prescribed portion of the reject portion is any portion of the reject portion. That is, in the 2 nd control, if the value of the dimension is the 2 nd value and any part of the reject part is separated by a predetermined distance or more from the center of the injection range of the air in the orthogonal direction of the 1 st nozzle, the air is injected from the 1 st nozzle among the plurality of nozzles and the nozzles adjacent to the 1 st nozzle. According to this aspect, the chance of ejecting air from 2 nozzles adjacent to each other increases as compared with the 7 th aspect. Therefore, the removal rate of the sorted material having the reject portion is improved. Therefore, a product with better quality can be obtained.

Drawings

Fig. 1 is a schematic diagram showing a schematic configuration of an optical sorter according to 1 embodiment of the present disclosure.

Fig. 2 is a flowchart showing a flow of the sorting process according to 1 embodiment.

Fig. 3 is a diagram showing 1 example of an image showing a defective portion.

Fig. 4 is a flowchart showing a flow of the nozzle determination processing according to 1 embodiment.

Fig. 5 is a schematic diagram showing a method of determining a nozzle corresponding to the size (dimension) of a defective portion, and shows a case where the value of the size of the defective portion is larger than a threshold value.

Fig. 6 is a schematic diagram showing a method of determining a nozzle corresponding to the size of the defective portion, and shows a case where the value of the size of the defective portion is equal to or smaller than a threshold value.

Fig. 7 is a flowchart showing a flow of the injection period determination processing of 1 embodiment.

Fig. 8 is a schematic view showing the effect of the optical sorter.

Fig. 9 is a schematic diagram showing the effect of the optical sorter.

Detailed Description

Fig. 1 is a schematic diagram showing a schematic configuration of an optical sorter (hereinafter, simply referred to as a sorter) 10 as 1 embodiment of the present disclosure. In the present embodiment, the classifier 10 is used to classify foreign matter (e.g., cobble, mud, glass flakes, etc.) and defective products (e.g., immature grains, colored grains, etc.) from rice grains (more specifically, brown rice or polished rice) as the objects to be classified 90. However, the objects to be sorted 90 are not limited to brown rice or polished rice, and may be any granular objects. For example, the objects to be sorted 90 may be rice, wheat, beans (soybean, chickpea, green soybean, etc.), resins (pellets, etc.), rubber sheets, etc.

As shown in fig. 1, the separator 10 includes: the light sources 30a, 30b, the optical sensors 40a, 40b, the sorting unit 50, the storage tank 71, the feeder 72, the chute 73, the reject discharge tank 74, the reject discharge tank 75, and the controller 80. The controller 80 controls the overall operation of the classifier 10. The controller 80 also functions as a detection unit 81 and an injection control unit 82. The function of the controller 80 may be realized by a CPU executing a predetermined program, may be realized by a dedicated circuit, or may be realized by a combination of these. The functions of the controller 80 may be realized by 1 device. For example, each function of the controller 80 may be realized by 1 CPU. Alternatively, the functions of the controller 80 may be distributed among at least 2 devices. The function of the controller 80 will be described later.

The accumulation tank 71 temporarily accumulates the sorted objects 90. The feeder 72 supplies the sorted objects 90 stored in the storage tank 71 to a chute 73 as an example of 1 sorted object transfer means. The objects 90 supplied to the chute 73 slide downward on the chute 73, and drop from the lower end of the chute 73. The chute 73 has a predetermined width that allows a plurality of objects 90 to be sorted to drop simultaneously. In the following description, a direction in which the objects 90 are transferred after falling from the chute 73 (in other words, a falling direction of the objects 90) is also referred to as a transfer direction D1. The direction perpendicular to the transfer direction D1 (in other words, the width direction of the chute 73) is also referred to as the perpendicular direction D2.

The light source 30a is disposed on one side (also referred to as front side) of the transfer path 95 of the object 90 (in other words, the falling trajectory of the object 90), and the light source 30b is disposed on the other side (also referred to as rear side) of the transfer path 95 of the object 90. The light sources 30a and 30b irradiate the light 31a and 31b, respectively, onto the objects 90 being sorted (i.e., the objects 90 sliding off the chute 73) being transferred on the transfer path 95. In the present embodiment, the light sources 30a and 30b are light source units each including a plurality of LEDs emitting red light, a plurality of LEDs emitting green light, and a plurality of LEDs emitting blue light. However, the specifications of the light sources 30a, 30b (for example, the number, the emission form, the wavelength regions of the lights 31a, 31b, and the like) are not particularly limited. For example, LEDs emitting near infrared light may be used as the light sources 30a and 30b instead of or in addition to the LEDs emitting visible light. In addition, one of the light sources 30a, 30b may be omitted.

The optical sensor 40a is disposed on the front side, and the optical sensor 40b is disposed on the rear side. The optical sensors 40a and 40b detect light irradiated from the light sources 30a and 30b and associated with the sorted object 90. Specifically, the front optical sensor 40a can detect the light 31a irradiated from the front light source 30a and reflected by the object 90 and the light 31b irradiated from the rear light source 30b and transmitted through the object 90. The rear optical sensor 40b can detect the light 31b irradiated from the rear light source 30b and reflected by the object to be sorted 90 and the light 31a irradiated from the front light source 30a and transmitted through the object to be sorted 90.

In the present embodiment, each of the optical sensors 40a and 40b is a color CCD sensor, and includes a plurality of light receiving elements arranged in a straight line. The plurality of light receiving elements are arranged in the orthogonal direction D2 (i.e., the width direction of the chute 73). Therefore, the optical sensors 40a and 40b can simultaneously capture a plurality of objects to be sorted 90 transferred over a predetermined width of the chute 73. The specifications of the optical sensors 40a and 40b are not particularly limited, and can be arbitrarily determined according to the specifications of the light sources 30a and 30b. In addition, one of the optical sensors 40a and 40b may be omitted.

The analog signals indicating the intensities of the detected lights, which are the outputs from the optical sensors 40a and 40b, are amplified by a predetermined gain by an AC/DC converter (not shown) and further converted into digital signals. The digital signal (in other words, a gradation value corresponding to an analog signal) is input to the controller 80. The controller 80 detects the reject portion of the sorted object 90 based on the detection result (i.e., image) of the inputted light. In the present embodiment, the reject portion includes a portion of rice having a color different from that of high-quality rice and the whole of foreign matter (i.e., a portion other than rice grains). However, the reject portion may be one of them. The colored rice is colored, and the colored rice is dead rice, damaged rice, or immature rice.

The sorting section 50 sorts the sorted objects 90 having the reject portion by injecting air toward the detected reject portion. Specifically, the sorting unit 50 includes a plurality of nozzles 51 and a number of solenoid valves 52 corresponding to the nozzles 51 (in the present embodiment, the number of the nozzles 51 is the same as that of the nozzles 51, but may be different from that of the nozzles 51). The plurality of nozzles 51 are arranged in the orthogonal direction D2 (i.e., the width direction of the chute 73).

The plurality of nozzles 51 are connected to a compressor (not shown) via a plurality of solenoid valves 52, respectively. By selectively opening the plurality of solenoid valves 52 in accordance with a control signal from the controller 80, the plurality of nozzles 51 selectively eject the air 53 toward the sorted objects 90 having the reject portions. More specifically, each of the plurality of nozzles 51 corresponds to a spraying responsible range in the orthogonal direction D2 with respect to each detection position of the sorted object 90. Each of the plurality of nozzles 51 ejects the air 53 when a predetermined portion (center in the present embodiment) of the defective portion is located within the corresponding ejection responsible range. Such control of the injection of the air 53 is performed as a process of the injection control section 82 of the controller 80. In this way, each of the plurality of nozzles 51 is allocated a range in the orthogonal direction D2 that is responsible for the ejection of the air 53 in relation to each detection position of the sorted object 90. In practice, the injection responsible range is defined by the position of the orthogonal direction D2 on the image input to the controller 80. Although described in detail later, the injection charge range is set to be variable in the present embodiment.

The sorted objects 90 having the reject portion are blown off by the air 53, and are separated from the falling rail (i.e., the transfer path 95) from the chute 73, and are guided to the reject discharge chute 75 (shown as the sorted objects 91 in fig. 1). On the other hand, air 53 is not injected to the sorted objects 90 (hereinafter, also referred to as "acceptable products") having no reject portion. Therefore, the acceptable product is guided to the acceptable product discharge groove 74 (indicated as the sorted product 92 in fig. 1) without changing the falling rail. Instead of the configuration in which the air 53 is sprayed toward the objects 90 after falling from the chute 73, the conveying path of the objects 90 may be changed by spraying the air 53 toward the objects 90 sliding on the chute 73. Further, as the sorted object transfer means, a conveyor may be used instead of the chute 73. In this case, air may be injected toward the objects to be sorted that fall from one end of the conveyor belt. Alternatively, air may be injected toward the objects being sorted while being conveyed on the conveyor belt.

The sorting process of the sorted objects 90 according to the present embodiment includes control of the air injection range with respect to the sorted objects 90 to be set to be variable. The following describes the sorting process in detail. Fig. 2 is a flowchart showing a flow of the sorting process performed by the controller 80. This process is repeatedly executed every time image data of a prescribed size is input to the controller 80. When the sorting process is started, first, the controller 80 detects a defective portion of the sorted object 90 based on the input image data as a process of the detection unit 81 (step S110). The detection of the defective portion is performed by comparing a threshold value predetermined to determine whether or not the defective portion is the defective portion with a gradation value of the image data. If the gradation value of each pixel constituting the image data is binarized based on whether or not the threshold value is equal to or less, the defective portion can be easily detected. For example, in the case of acquiring image data of each of RGB colors, binarization may be performed using threshold values set for each color. In this case, the pixels indicating the defective portions are an aggregate of pixels determined as defective portions with respect to any one color.

Next, the controller 80 determines whether or not the defective portion is detected in step S110 (step S120). As a result of the judgment, if no reject portion is detected (no in step S120), the controller 80 ends the sorting process.

On the other hand, if a defective portion is detected (step S120: yes), the controller 80 advances the process to step S130. In the case where a plurality of reject sections are detected in step S110, the processing of step S130 and thereafter is performed for each reject section. On an image, a single defective portion is constituted by a plurality of continuous pixels having gradation values representing the defective portion, and a group of discontinuous pixels having gradation values representing the defective portion is determined to be mutually different defective portions.

In step S130, the controller 80 detects the value of the size of the detected defective portion as a process of the detecting unit 81. The dimensions refer to any one of or any combination of the area, length, and width of the reject fraction. In the present embodiment, the size is the area of the reject portion. Fig. 3 shows 1 example of an image showing a single reject portion. In fig. 3, each grid represents pixels constituting image data. For convenience of explanation, the X axis is defined parallel to the orthogonal direction D2, and the Y axis is defined parallel to the transfer direction D1. Pixels displaying the reject fraction are indicated by section lines (single section lines or cross section lines). In the example shown in fig. 3, the reject portion is constituted by 23 pixels, and the controller 80 detects the value of the area of the reject portion as 23.

In an alternative embodiment, when the size is defined as a combination of the length L1 and the width W1, the controller 80 may detect the value 5 of the length L1 and the value 7 of the width W1 based on the reject portion shown in fig. 3. In this alternative embodiment, the length L1 is determined based on the maximum range of the defective portion in the transfer direction D1, and the width W1 is determined based on the maximum range of the defective portion in the orthogonal direction D2.

Next, the controller 80 detects the center of the defective portion as a process of the detecting unit 81 (step S140). In the present embodiment, the controller 80 detects a pixel located at the center of the pixel position of the predetermined length L1 and at the center of the pixel position of the predetermined width W1 as the center of the defective portion. In the example of fig. 3, since the length L1 is defined by the pixel positions Y2 to Y6 and the width W1 is defined by the pixel positions X2 to X8, the pixel (indicated by cross hatching) of (X, Y) = (X5, Y4) is detected as the center. When the number of pixel positions of the predetermined length L1 is even, the center may be determined based on any of 2 pixel positions located at the center of the pixel positions. The same applies to the case where the number of pixel positions of the predetermined width W1 is even.

The detection of the center is not limited to this method, and can be performed by any known method. For example, the controller 80 may remove pixels constituting the outline of the defective portion from the pixel group constituting the defective portion one by one clockwise, and detect the last remaining pixel as the center of the defective portion.

Next, the controller 80 executes a nozzle determination process as a process of the injection control section 82 (step S150). The nozzle determination process is a process of determining whether to use any nozzle 51 among the plurality of nozzles 51 to eject the air 53 in order to sort the sorted objects 90 having the reject portion. The nozzle determination process is performed to determine control of the air injection range in the orthogonal direction D2 to the objects to be sorted 90 (hereinafter, also referred to as "control 2") on the condition of one of the values of the sizes of the reject portions. The 2 nd control is 1 of the above-described injection range controls. In the present embodiment, step S150 is performed with reference to the center of the reject portion. The nozzle determination process will be described in detail below.

Fig. 4 is a flowchart showing a flow of the nozzle determination processing according to 1 embodiment. In this process, the controller 80 first determines whether or not the value of the size detected in step S130 is equal to or smaller than a threshold value (step S151). In the present embodiment, since the size is defined as an area, a threshold value is set in advance with respect to the value of the area. A value of the dimension below the threshold means that the reject fraction is small to a prescribed extent. The threshold value may be set to, for example, about 1/3 of the standard size of the sorted object 90. In this case, in step S151, it is substantially determined whether or not the reject portion is about 1/3 or less of the standard size of the sorted object 90.

As a result of the determination, if the value of the size is greater than the threshold (step S151: no), the controller 80 sets the injection responsible range of the non-overlapping region (step S152). The "non-overlapping region" means that the injection responsible ranges assigned to the respective nozzles 51 do not overlap with each other. Fig. 5 shows 1 example of the correspondence relationship between the nozzle 51 and the injection charge range in the case where the injection charge range is set as described above. In fig. 5, for convenience of explanation, only 3 nozzles 51 in the middle of the arrangement among the plurality of nozzles 51 arranged in the orthogonal direction D2 are shown as nozzles 51a to 51c. The electromagnetic valves 52a to 52c are connected to the nozzles 51a to 51c, respectively. In fig. 5, the grids represent pixels constituting the image 85 inputted to the controller 80, respectively. As shown in fig. 5, the nozzles 51a to 51c correspond to the injection responsible ranges 86a to 86c, respectively. The injection responsible ranges 86a to 86c are set so as not to overlap each other. Although not shown in the drawings, in reality, all the pixel positions on the image 85 aligned in the orthogonal direction D2 (in other words, all the detection positions of the sorted objects 90 in the orthogonal direction D2) belong to any one of the ejection responsible ranges corresponding to the plurality of nozzles 51.

On the other hand, if the value of the dimension is equal to or less than the threshold value (yes in step S151), the controller 80 sets the injection charge range having the overlap region (step S153). The "overlapped region" means that any adjacent 2 of the plurality of injection responsible regions allocated to each nozzle 51 are partially overlapped. Fig. 6 shows 1 example of the correspondence relationship between the nozzle 51 and the injection charge range in the case where the injection charge range is set as described above. In this example, only 3 nozzles 51a to 51c are shown, as in fig. 5. In fig. 6, the correspondence between the nozzles 51a and 51c and the injection charge ranges 86a and 86c corresponding thereto is indicated by a one-dot chain line, and the correspondence between the nozzle 51b and the injection charge range 86b corresponding thereto is indicated by a broken line. As shown in fig. 6, the 2 injection responsible regions 86a, 86b corresponding to the adjacent 2 nozzles 51a, 51b overlap each other in the overlap region 87 a. Similarly, the 2 injection responsible ranges 86b and 86c corresponding to the adjacent 2 nozzles 51b and 51c overlap each other in the overlap region 87b.

In an alternative embodiment, when the size is defined as a combination of the length L1 and the width W1, the threshold value used in step S151 is set in advance for both the length L1 and the width W1. In this case, in step S151, if at least one of the length L1 and the width W1 is equal to or less than the 1 st threshold and equal to or less than the 2 nd threshold is satisfied, the process may be advanced to step S153. In still another alternative embodiment, in step S151, the process may be advanced to step S153 only when the length L1 is equal to or less than the 1 st threshold and the width W1 is equal to or less than the 2 nd threshold. Although it is not known in which direction the objects 90 (and hence the reject fraction) are transferred, the size of the reject fraction can be accurately grasped regardless of the direction of the objects 90 during transfer, by defining the dimensions in advance with both the length L1 and the width W1.

Next, the controller 80 determines, from among the injection charge ranges set in step S152 or step S153, the nozzle 51 corresponding to the injection charge range to which the center of the reject fraction detected in step S140 belongs, as a nozzle (hereinafter also referred to as a nozzle to be ejected) to which the air 53 should be ejected in order to sort the sorted objects 90 having the reject fraction (step S154). In other words, the nozzle to be ejected is determined with reference to the center of the defective portion.

In the example shown in fig. 5, the center 84 of the reject portion exists in the injection responsible range 86b. Therefore, the nozzle 51b corresponding to the injection charge range 86b is determined as the nozzle to be injected. In the example shown in fig. 6, the center 84 of the reject fraction is located within the overlap region 87 a. That is, the center 84 of the reject fraction is located within the injection responsible range 86a, and is also located within the injection responsible range 86b. Accordingly, the nozzles 51a and 51b corresponding to the injection responsible ranges 86a and 86b are determined as nozzles to be injected.

In this way, in the nozzle determination processing, when the processing proceeds from step S152 to step S154, no matter where the position of the center in the orthogonal direction D2 is, the single nozzle 51 is determined in step S154. On the other hand, in the case where the process proceeds from step S153 to step S154, when the center is located in the overlapping region in the orthogonal direction D2, 2 nozzles 51 are determined in step S154, and when the center is not located in the overlapping region in the orthogonal direction D2, a single nozzle 51 is determined in step S154.

According to the above-described nozzle determination processing, when the value of the size is larger than the threshold value, the single nozzle 51 is determined as the nozzle to be ejected, and when the value of the size is equal to or smaller than the threshold value and the defective portion is separated from the center of the ejection range of the air 53 in the orthogonal direction D2 of any one of the nozzles 51 by a predetermined distance or more in the orthogonal direction D2, the single nozzle 51 and the adjacent nozzle (more specifically, one of the nozzles 51 adjacent to the single nozzle 51, which is close to the center of the defective portion in the orthogonal direction D2) are determined as the nozzle to be ejected.

Here, the description will be returned to the sorting process of fig. 2. When the nozzle determination process is completed, the controller 80 executes an injection period determination process as a process of the injection control unit 82 (step S160). The injection period determination process refers to a process of determining the injection period of the air 53 (i.e., the time to continue the injection). The injection period determination process is performed to perform control (hereinafter, also referred to as "1 st control") for determining the air injection range in the transfer direction D1 with respect to the objects to be sorted 90 on the condition of the extent of the size of the reject portion. The air injection range here refers to a range in a coordinate system that moves together with the object 90 being sorted while being transferred. The 1 st control is 1 of the above-described injection range controls. The injection period determination process will be described in detail below.

Fig. 7 is a flowchart showing a flow of the injection period determination processing of 1 embodiment. In this process, the controller 80 first determines whether or not the value of the size detected in step S130 is equal to or smaller than a threshold value (step S161). The threshold used in step S161 is typically the same as the threshold used in step S151. However, the two may be different.

As a result of the determination, if the value of the size is greater than the threshold (step S161: no), the controller 80 sets the injection period to T1 (step S162). On the other hand, if the value of the size is equal to or less than the threshold value (step S161: yes), the controller 80 sets the injection period to a period T2 longer than the period T1 (step S163).

Here, the description will be returned to the sorting process of fig. 2. When the injection period determination process is completed, the controller 80 controls the solenoid valve 52 so that the air 53 is injected from the nozzle 51 (nozzle to be injected) determined in step S150 in the injection period determined in step S160 (step S170) as a process of the injection control unit 82, and the sorting process is completed. In the case where the number of nozzles to be ejected is plural, air 53 is ejected simultaneously from these nozzles to be ejected.

In the present embodiment, the injection delay time is set with reference to the center of the defective portion with respect to the injection of the air 53 in step S170. Specifically, the injection delay time is set so that the timing of injecting air 53 with respect to the center of the defective portion is located at the center of the injection period. In other words, regardless of which of the periods T1 and T2 is set as the injection period, in this case, the timings of starting and ending the injection period are set so as to be located at the center of the injection period with respect to the timing of the center injection air 53 of the defective portion. In this way, the air 53 can be reliably brought into contact with the sorted object 90 having the reject portion, regardless of which side in the transfer direction D1 of the sorted object 90 the reject portion is biased.

According to the above-described separator 10, the outline or center of the sorted object 90 is not detected, but the injection range of the air 53 is appropriately controlled based on the value of the size of the reject portion of the sorted object 90, and the air 53 is injected toward the reject portion. Since the outline or center of the object 90 to be sorted does not need to be detected, the load of signal processing is relatively small, and the processing can be performed at high speed. In addition, even if the plurality of objects to be sorted 90 are transferred in a contact state, the sorting accuracy is not lowered.

When the value of the size of the reject portion is equal to or smaller than the threshold value, the range of contact between the air 53 in the transfer direction D1 and the objects 90 is widened by the 1 st control. When the value of the dimension of the reject fraction is equal to or less than the threshold value and the reject fraction is separated from the center of the injection range of the air 53 in the orthogonal direction D2 of any of the nozzles 51 by a predetermined distance or more in the orthogonal direction D2, the contact range between the air 53 in the orthogonal direction D2 and the objects 90 to be sorted is widened by the 2 nd control. Therefore, a small contact phenomenon is less likely to occur, and the sorting accuracy can be improved.

Fig. 8 and 9 are schematic views showing the effects of the classifier 10. In fig. 8 and 9, the grids represent pixels constituting the image 85, respectively. The hatched portion represents the reject portion and the black square represents the center of the reject portion. The left side of fig. 8 and 9 shows a case where the value of the size of the defective portion is greater than the threshold value, and the right side shows a case where the value of the size of the defective portion is equal to or less than the threshold value. In fig. 8 and 9, the arrow range indicates the injection range of 1 nozzle 51.

As shown in the left side of fig. 8, when the reject portion is large to some extent and the reject portion is located near the end of one side of the sorted object 90 in the transfer direction D1, the reject portion overlaps the sorted object 90 by a sufficient extent with respect to the injection range A1 of the sorted object 90 (which is not expanded by the 1 st control), and therefore, a small contact phenomenon is difficult to occur. In addition, as shown in the right side of fig. 8, when the reject portion is small and the reject portion is located near the end of the one side of the sorted object 90 in the transfer direction D1, the ejection range a11 in the conventional sorting process is small compared with the repeated portion of the sorted object 90, so that there is a possibility that a small contact phenomenon occurs. On the other hand, according to the present embodiment, the injection period of the air 53 is set longer than in the case of the left side of fig. 8 by the 1 st control, and as a result, the injection range A2 that is expanded in the direction of the transfer direction D1 can be set. Therefore, the contact range between the object 90 and the air 53 is widened, and the occurrence of the small contact phenomenon can be suppressed.

In addition, as shown in the left side of fig. 9, in the case where the reject portion is large to some extent and the reject portion is located near the end of the one side of the sorted object 90 in the orthogonal direction D2, the small contact phenomenon is hardly generated because the reject portion is repeated with the sorted object 90 by a sufficient magnitude with respect to the ejection range A1 of the sorted object 90 (which is not expanded by the 1 st control or the 2 nd control). In addition, as shown on the right side of fig. 9, when the reject portion is small and the reject portion is located near the end of the sorted object 90 on one side in the orthogonal direction D2, the overlap between the ejection range a11 and the sorted object 90 in the conventional sorting process is very small, and therefore, there is a possibility that a small contact phenomenon occurs. On the other hand, according to the present embodiment, the injection period of the air 53 is set longer than the case of the left side of fig. 9 by the 1 st control, and the number of the injection nozzles to be injected is increased to 2 by the 2 nd control, so that the injection range A2 which is expanded in both the transfer direction D1 and the orthogonal direction D2 can be set. Therefore, the contact range between the object 90 and the air 53 is widened, and the occurrence of the small contact phenomenon can be suppressed.

In the alternative embodiment, in the 1 st control, when the value of the size of the reject fraction is equal to or smaller than the threshold value (yes in step S161 in fig. 7), the injection period may be set to be variable so that the injection period becomes longer stepwise as the value of the size becomes smaller. In this case, the injection period may be set to vary linearly or stepwise.

In the alternative embodiment, in the 2 nd control, when the value of the size of the defective portion is equal to or smaller than the threshold value (yes in step S151 in fig. 4), the overlap region may be set to be variable so that the smaller the value of the size is, the larger the overlap region (in the example of fig. 6, the overlap regions 87a and 87 b) stepwise becomes. In this case, the overlapping region may be set to change linearly or stepwise.

In the alternative embodiment, the 1 st control is not limited to the configuration performed with reference to the center of the defective portion. In the 1 st control, the timings of starting and ending the injection period may be set so as to be located at the center of the injection period with respect to the timing of injecting the air 53 at the predetermined portion among the defective portions. In the above embodiment, the predetermined portion that serves as the reference for the 1 st control is set as the center of the reject portion, but in an alternative embodiment, it may be arbitrarily selected and set in advance among the reject portions.

In the alternative embodiment, the 2 nd control is not limited to the configuration performed with reference to the center of the defective portion. In the control of fig. 2, instead of step S154 (see fig. 4), the nozzle 51 corresponding to the injection charge range 86 to which a predetermined portion (in other words, a predetermined pixel) among the defective portions belongs may be determined as the nozzle to be injected. The predetermined portion may be set as any portion among the reject portions. That is, the nozzles 51 corresponding to all of the injection responsible ranges 86 to which any of the defective portions belongs may be determined as nozzles to be injected. In this case, for example, when the entire defective portion exists throughout the 2 injection responsible regions 86, the air 53 is injected from the 2 nozzles 51 corresponding to the 2 injection responsible regions 86 regardless of whether or not the center of the defective portion exists in the overlap region. According to this alternative embodiment, the chances of ejecting air from 2 nozzles 51 adjacent to each other increases, and the removal rate of the sorted objects 90 having defective portions increases. Therefore, a product of better quality can be obtained. The controller 80 may be configured to switch the control mode between the configuration of this alternative embodiment and the configuration shown in fig. 4 depending on whether the user prioritizes the yield or the quality. However, the predetermined portion serving as the reference of the 2 nd control may be arbitrarily selected and set in advance from among the reject portions.

In an alternative embodiment, in the case where the pixel indicating the defective portion is present at the end of the transfer direction D1 in the image in the detection of the size of the defective portion (step S130 in fig. 2), the controller 80 may determine the continuity of the defective portion between the plurality of images by further referring to the image inputted in the first 1 or the last 1 of the images. In the case of continuity, the outline of the defective portion may be determined over a plurality of images, and the size of the defective portion may be detected based on the outline.

The embodiments of the present disclosure have been described above, but the embodiments are for easy understanding of the present invention and are not limited to the present invention. The present invention can be modified and improved without departing from the gist thereof, and equivalents thereof are included in the present invention. In addition, any combination or any omission of the respective constituent elements described in the claims and the description may be made in a range in which at least a part of the above-described problems can be solved or a range in which at least a part of the effects can be obtained.

For example, the above-described flowchart is only 1 example, and each process constituting the flowchart can be changed in the process order or in equivalent processes without departing from the scope of the gist of the present invention.

In addition, only one of the above-described control 1 and control 2 may be used alone.

Description of the reference numerals

Optical sorter; 30a, 30b. 31a, 31 b..light; optical sensors; sorting section; 51. 51a, 51b, 51c. 52. 52a, 52b, 52 c..solenoid valves; 53. air; 71. the tank; 72. feeder; 73. chute; 74. the acceptable product discharge tank; 75. reject discharge tank; a controller; 81. a detection unit; 82. an injection control unit; 84. center; 85. the image; 86a, 86b, 86c. 87a, 87 b..overlapping area; 90. 91, 92. the sorted material; 95. transfer path.

Claims (7)

1. An optical sorter comprising:

a light source configured to irradiate light to the conveyed objects to be sorted;

an optical sensor configured to detect light irradiated from the light source and associated with the sorted material;

a detection unit configured to detect a defective portion of the sorted object based on a signal acquired by the optical sensor;

a sorting section configured to sort the sorted objects having the reject portion by injecting air toward the reject portion; and

an injection control unit configured to control injection of the air,

the injection control unit is configured to execute injection range control in which the air is injected into the 1 st injection range with respect to the sorted object when the value of the size of the reject fraction is the 1 st value, and the air is injected into the 2 nd injection range with respect to the sorted object when a predetermined condition is satisfied,

the 2 nd injection range is wider than the 1 st injection range,

the prescribed condition includes that the value of the size is a2 nd value smaller than the 1 st value.

2. The optical sorter of claim 1 wherein,

the injection range control includes a1 st control in which, when the value of the size is the 2 nd value, the injection period of the air is set to a longer time than when the value of the size is the 1 st value.

3. An optical sorter as in claim 1 or 2 wherein,

the sorting section includes a plurality of nozzles that can selectively eject the air and are arranged in a direction orthogonal to a transfer direction of the objects to be sorted, that is, in an orthogonal direction,

the injection range control includes a2 nd control in which, when the value of the size is the 1 st value, the air is injected from 1 nozzle among the plurality of nozzles, and when the value of the size is the 2 nd value and a predetermined portion of the defective portion is separated from a center of an injection range of the air in the orthogonal direction of the 1 nozzle by a predetermined distance or more in the orthogonal direction, the air is injected from the 1 nozzle among the plurality of nozzles and a nozzle adjacent to the 1 nozzle.

4. The optical sorter of claim 3 wherein,

each of the plurality of nozzles variably corresponds to a spray responsibility range in the orthogonal direction with respect to each detection position of the sorted object,

the injection control section is further configured to control the injection of the air so that each of the plurality of nozzles injects the air when the prescribed portion of the reject portion is within a corresponding injection charge range,

a plurality of injection responsible ranges corresponding to the plurality of nozzles, respectively, are set to be non-overlapping with each other when the value of the size is the 1 st value, and are set to be partially overlapping with each other when the value of the size is the 2 nd value, of any adjoining 2 injection responsible ranges among the plurality of injection responsible ranges,

and ejecting the air from 2 nozzles corresponding to the adjacent 2 injection responsible ranges when the predetermined portion of the reject portion is located within an overlapping region of the adjacent 2 injection responsible ranges.

5. The optical sorter as in any of claims 1 to 4 wherein,

the detecting section is further configured to detect a center of the reject portion,

the injection control unit is further configured to perform the injection range control with reference to the center of the defective portion.

6. An optical sorter as in claim 5 when dependent on claim 2 wherein,

the 1 st control includes setting timings of starting and ending the injection period so that a timing of injecting the air to the center of the defective portion is located at the center of the injection period.

7. An optical sorter as in claim 5 as dependent on claim 3 or claim 6 as dependent on claim 3,

the prescribed portion of the reject portion is the center of the reject portion.