Method for determining the presence or absence of a part and image processing system

Abstract

This component presence / absence determination method includes: extracting a plurality of types of feature quantities from each of a component present image and a component absent image for every feeder or every component type; selecting, on the basis of the extracted plurality of types of feature quantities, one or more types of feature quantities effective for determination, from among the plurality of types of feature quantities; and storing, in advance, the selected types of feature quantities as effective feature quantities. The method also includes, to determine whether a target component is present in a cavity of a target feeder, which is a determination target among the plurality of feeders: acquiring an image of a tape of the target feeder; extracting, from the acquired image, an effective feature quantity of a type selected for the target feeder or a component type of the target component; and determining whether the target component is present in the cavity of the target feeder on the basis of the extracted effective feature quantity and the effective feature quantities corresponding respectively to the component present image and the component absent image stored in advance.

Description

Technical Field

[0001] This specification discloses a method for determining the presence or absence of components and an image processing system.

Background Art

[0002] Conventionally, in a feeder that supplies a tape provided with a plurality of cavities for accommodating components, it is known to determine whether there is a component in the cavity from an image of the tape. For example, in Patent Document 1, in an image processing apparatus that processes an image of a tape, the luminance in a predetermined range including the cavity of the tape image is extracted, a value indicating the variation in the extracted luminance is obtained as a feature amount, and the obtained feature amount is compared with a threshold value to determine whether there is a component in the cavity. Further, Patent Document 1 discloses that a threshold value is determined for each type of tape or each type of component, the type of the tape or the type of the component to be processed is obtained, and the presence or absence of the component in the cavity is determined using the threshold value corresponding to the type. Furthermore, Patent Document 1 discloses that a threshold value is determined by machine learning using a plurality of types of feature amounts (five feature amounts of maximum luminance, minimum luminance, average luminance, contrast, and luminance variation) including luminance variance and standard deviation, and the presence or absence of the component in the cavity is determined based on the plurality of types of feature amounts extracted from the image of the tape to be processed and the threshold value.

Prior Art Documents

Patent Documents

[0003]

Patent Document 1

Summary of the Invention

Problems to be Solved by the Invention

[0004] Feeders can have individual differences due to factors such as differences between production lots or wear and tear from prolonged use, meaning that the effective feature quantities for detection may differ from one feeder to another. In this case, the method described in Patent Document 1 may, depending on the individual differences in the feeder, extract inappropriate feature quantities from the tape image, resulting in a misjudgment of the presence or absence of a component. Similar problems can occur depending on the type of component.

[0005] The primary purpose of this disclosure is to determine the presence or absence of components within a cavity from an image of a tape with good accuracy, regardless of the feeder or component type. [Means for solving the problem]

[0006] This disclosure employs the following means to achieve the primary objectives described above.

[0007] The method for determining the presence or absence of a component in this disclosure is: A method for determining whether or not a component is present in a cavity of a tape, in a feeder that supplies a tape having multiple cavities for housing components, wherein the method determines from an image of the tape whether or not a component is present in the cavity. For each feeder or part type, multiple types of feature quantities are extracted from the image with a part in the cavity and the image without a part in the cavity, and based on the extracted multiple types of feature quantities, one or more types of feature quantities that are effective for the determination are selected from among the multiple types of feature quantities, and the selected types of feature quantities are stored in advance as effective feature quantities. In determining whether or not a target component is present in the cavity of a target feeder among the plurality of feeders, an image of the tape of the target feeder is acquired, effective feature quantities of a selected type for the target feeder or the type of component of the target component are extracted from the acquired image, and a determination is made as to whether or not the target component is present in the cavity of the target feeder based on the extracted effective feature quantities and the corresponding effective feature quantities of the image with the component and the image without the component, which have been stored in advance. This is the gist of it.

[0008] In the part presence / absence determination method of this disclosure, for each feeder or part type, one or more types of feature quantities effective for determination are selected from multiple types of feature quantities based on multiple types of feature quantities extracted from images with and without parts. Then, when determining whether or not a target part is present in the cavity of the target feeder, the determination is made using the type of feature quantities selected for the target feeder or the part type of the target part. As a result, regardless of the feeder or part type, the presence or absence of a part in the cavity can be determined from the tape image with good accuracy.

[0009] Furthermore, the image processing system disclosed herein can achieve the same effects as the part presence / absence determination method disclosed herein. [Brief explanation of the drawing]

[0010] [Figure 1] This is a schematic diagram of the component mounting machine. [Figure 2] This is a close-up view of the mark camera and feeder. [Figure 3] This is a schematic diagram of the Mark Camera system. [Figure 4] This is a block diagram showing the electrical connection relationship between a component mounting machine and a control device. [Figure 5] This flowchart shows an example of pre-processing performed by the image processing unit. [Figure 6] This is an explanatory diagram showing the minimum brightness under R-light side emission for images with and without components. [Figure 7] This is an explanatory diagram showing the contrast between images with and without components when all B-lights are illuminated. [Figure 8] This is an explanatory diagram showing the maximum brightness under G-light emission for images with and without components. [Figure 9] This is an explanatory diagram showing the average brightness under R-light emission for images with and without components. [Figure 10] This is an explanatory diagram showing an example of information used for determining feeder types. [Figure 11] This flowchart shows an example of a component presence / absence determination process performed by the image processing unit.

Best Mode for Carrying Out the Invention

[0011] Next, embodiments for implementing the present disclosure will be described with reference to the drawings.

[0012] FIG. 1 is a schematic configuration diagram of the component mounter 10. FIG. 2 is a partially enlarged view of the mark camera 20 and the feeder 30. FIG. 3 is a schematic configuration diagram of the mark camera 20. FIG. 4 is a block diagram showing the electrical connection relationship between the component mounter 10 and the management device 60. In FIG. 1, the left-right direction is the X-axis direction, the front (near) - rear (far) direction is the Y-axis direction substantially orthogonal to the X-axis direction, and the up-down direction is the Z-axis direction substantially orthogonal to the X-axis direction and the Y-axis direction (horizontal plane).

[0013] The component mounter 10 takes out components from the feeder 30 and mounts them on the substrate S. As shown in FIG. 1, it includes a base 11, a substrate transfer device 12, a mounting head 14, a head moving device 16, a parts camera 18, a mark camera 20, a feeder 30, and a control device 40 (see FIG. 4). A plurality of component mounters 10 are arranged side by side in the substrate transfer direction to form a component mounting line. The component mounting line is managed by a management device 50 (see FIG. 4).

[0014] The substrate transfer device 12 is installed on the base 11. The substrate transfer device 12 has a pair of conveyor belts provided at intervals in the front-rear (Y-axis direction) and spanning in the left-right (X-axis direction). The substrate S is transferred from left to right in FIG. 1 by driving the conveyor belts.

[0015] The feeder 30 is attached to the feeder base installed on the base 11 so as to be arranged in the left - right direction (X - axis direction). The feeder 30 includes a reel around which a tape 32 for accommodating components is wound. As shown in FIG. 2, the tape 32 has a plurality of cavities 34 and sprocket holes 33 formed at equal intervals along its longitudinal direction. Components are accommodated in each cavity 34. These components are protected by a film covering the surface of the tape 32. The components in the cavity 34 are exposed at the component supply position by peeling off the film in front of the component supply position and are picked up (adsorbed) by the mounting head 14.

[0016] The mounting head 14, although not shown, includes a holder for holding the suction nozzle 15 and a lifting device for raising and lowering the holder. The suction nozzle 15 has a suction port at its tip and can adsorb components by the negative pressure supplied from a negative pressure source (not shown) to the suction port. The mounting head 14 may be a head having a single suction nozzle 15, or may be a rotary head having a plurality of suction nozzles 15 arranged at equal intervals along the outer periphery of a cylindrical head body. Also, as a member for holding components, instead of the suction nozzle 15, a mechanical chuck or an electromagnetic chuck may be used.

[0017] The head moving device 16 moves the mounting head 14 in the front - rear, left - right (XY - axis directions). As shown in FIG. 1, the head moving device 16 includes an X - axis slider 16a and a Y - axis slider 16b. The X - axis slider 16a is supported by an X - axis guide rail provided so as to extend in the left - right (X - axis direction) with respect to the Y - axis slider 16b and can move left and right by the drive of an X - axis motor. The Y - axis slider 16b is supported by a Y - axis guide rail provided so as to extend in the front - rear (Y - axis direction) and can move left and right by the drive of a Y - axis motor. The mounting head 14 is attached to the X - axis slider 16a. Therefore, the mounting head 14 can move along the XY plane (horizontal plane) by driving and controlling the head moving device 16 (X - axis slider 16a and Y - axis slider 16b).

[0018] The parts camera 18 is mounted on the base 11. When a part that has been picked up by the suction nozzle 15 passes above the parts camera 18, the parts camera 18 takes an image of the part from below and outputs the captured image to the control device 40 (see Figure 4).

[0019] The mark camera 20 is mounted on the X-axis slider 16a (or mounting head 14) and moves together with the mounting head 14 in the forward, backward, left, and right directions (XY axis directions) by the head moving device 16. The mark camera 20 captures an image of the object to be imaged from above and outputs the captured image to the control device 40 (see Figure 4). Examples of objects to be imaged include marks attached to the substrate S, tapes 32 (cavities 34) fed out by the feeder 30, and components after they have been mounted on the substrate S.

[0020] As shown in Figure 3, the Mark Camera 20 comprises an illumination unit 21, a lens 25, and an imaging unit 26.

[0021] The illumination unit 21 comprises a side illumination unit 22 and an incident illumination unit (coaxial incident illumination unit) 23. The side illumination unit 22 illuminates the object at an oblique angle. The side illumination unit 22 has multiple light sources of different colors, for example, a red LED 22r that emits monochromatic R (red) light, a green LED 22g that emits monochromatic G (green) light, and a blue LED 22b that emits monochromatic B (blue) light. Multiple LEDs 22r, 22g, and 22b are arranged in a ring shape around the lens 25 when viewed from above. The incident illumination unit 23 illuminates the object from the same direction as the optical axis of the lens 25. The incident illumination unit 23 has a half mirror 24 positioned at a 45-degree angle to the optical axis of the lens 25, and a light source that illuminates the half mirror 24 in a direction perpendicular to the optical axis of the lens 25 (horizontal direction). The light source of the incident illumination unit 23 includes multiple light sources of different colors, for example, a red LED 23r that emits monochromatic R (red) light, a green LED 23g that emits monochromatic G (green) light, and a blue LED 23b that emits monochromatic B (blue) light.

[0022] The illumination unit 21 has three light irradiation patterns: a side-emitting irradiation pattern in which light is emitted only from the side-emitting illumination unit 22; an incident irradiation pattern in which light is emitted only from the incident illumination unit 23; and a full illumination irradiation pattern in which light is emitted from both the side-emitting illumination unit 22 and the incident illumination unit 23. Furthermore, the illumination unit 21 has a red light irradiation pattern in which R light is emitted, a green light irradiation pattern in which G light is emitted, and a blue light irradiation pattern in which B color light is emitted, for each of the side-emitting, incident, and full illumination patterns.

[0023] The imaging unit 26 includes a monochrome image sensor (e.g., a monochrome CCD) that generates a monochromatic image based on the received light. The imaging unit 26 receives light emitted from the side illumination unit 22 and the incident illumination unit 23, which is reflected by the object, via the half mirror 24, and generates an image.

[0024] The control device 40 includes a CPU 42, ROM 44, RAM 46, a storage unit 48 such as a hard disk or solid-state drive, input / output ports (not shown), and communication ports. It controls the various drive units of the component mounting machine 10 and processes images captured by the parts camera 18 and mark camera 20. The control device 40 outputs various control signals to the feeder 30, board transport device 12, mounting head 14, head moving device 16, parts camera 18, mark camera 20, etc. The control device 40 also receives detection signals from various sensors, including a position sensor that detects the position of the mounting head 14, and receives image signals from the parts camera 18 and mark camera 20. Furthermore, the control device 40 is connected to a management device 50 that manages the component mounting line, including the component mounting machine 10, via a communication network, enabling bidirectional communication and the exchange of data and control signals between them.

[0025] The management device 50 is a general-purpose computer and, as shown in Figure 4, comprises a management control unit 52 consisting of a CPU, ROM, RAM, etc., an input device 54 such as a keyboard or mouse, a display 56, and a storage unit 58 such as a hard disk or solid-state drive. The storage unit 58 stores the production schedule for the circuit board S. The production schedule specifies which components to mount on which circuit boards in the component mounting machine 10 and how many circuit boards with components mounted in that manner will be produced. The management device 50 instructs the control device 40 of the component mounting machine 10 to produce circuit boards S with components mounted according to the production schedule.

[0026] When the CPU 42 of the control device 40 receives a production instruction from the management device 50, it first determines whether the pre-processing for obtaining the feature quantities of the image with parts and the image without parts, which are necessary for executing the part presence / absence determination process described later, has been performed for the feeder 30 that supplies the target parts to be mounted. The image with parts is an image generated by capturing an image of the tape 32 with parts contained in the cavity 34 using the mark camera 20. The image without parts is an image generated by capturing an image of the tape 32 with no parts contained in the cavity 34 using the mark camera 20. Feature quantities are quantities that characterize the image and are used to determine the presence or absence of parts. The pre-processing for determination is performed for each feeder 30.

[0027] The control device 40 performs pre-determination processing for the target feeder 30 if it has not already been performed. On the other hand, if the pre-determination processing has already been performed, the CPU 42 performs a component presence determination process to determine whether or not there are components in the cavity 34 of the tape 32 sent out from the target feeder 30. If the CPU 42 determines that there are no components in the cavity 34, it determines that the target feeder 30 has run out of components and outputs an error. The feeder 30 that has run out of components is collected and replaced by an operator or an automated replacement robot (not shown).

[0028] Meanwhile, when the CPU 42 determines that there is a component in the cavity 34, it performs a suction operation to pick up the component supplied from the target feeder 30 onto the suction nozzle 15. Specifically, the control device 40 controls the head moving device 16 to move the mounting head 14 directly above the component supply position of the target feeder 30. Next, it controls a lifting device (not shown) to lower the suction nozzle 15 and controls a negative pressure source (not shown) to supply negative pressure to the suction nozzle 15. As a result, the component is picked up by the tip of the suction nozzle 15. Next, the control device 40 raises the suction nozzle 15 and controls the head moving device 16 to move the suction nozzle 15, which has picked up the component, above the target mounting position on the substrate S. Then, the control device 40 lowers the suction nozzle 15 and controls a positive pressure source (not shown) to supply atmospheric pressure to the suction nozzle 15. As a result, the component that was picked up by the suction nozzle 15 separates from the suction nozzle 15 and is mounted on the substrate S.

[0029] Next, we will explain the details of the pre-determination processing. Figure 5 is a flowchart showing an example of pre-determination processing performed by the CPU 42 of the control device 40.

[0030] In the pre-determination processing, the CPU 42 first determines whether or not the suction operation has not yet been performed on the target feeder 30 (S100). If the CPU 42 determines that the suction operation has not yet been performed, the CPU 42 controls the mark camera 20 to capture an image of the tape 32 (cavity 34) before the tape is fed out of the target feeder 30 and the part is removed by the suction operation, as an image with a part (S110). The image with a part is captured multiple times while changing the imaging conditions (side illumination, incident illumination, and illumination patterns with combinations of R, G, and B colored light). Since the pre-determination processing is performed after production has started and before the first suction operation is performed on the target feeder 30, it can be determined that a part is contained in the cavity 34 of the tape 32 fed out of the target feeder 30, that is, the image of the tape 32 can be determined to be an image with a part. Note that confirmation of whether or not an image has a part may be performed visually by an operator.

[0031] Next, the CPU 42 determines whether or not the suction operation has been performed (S120). If the CPU 42 determines that the suction operation has been performed, the mark camera 20 controls the mark camera 20 to capture an image of the tape 32 (cavity 34) after it has been sent to the target feeder and the part has been picked up by the suction operation, as an image of the part-free image (S130). The image of the part-free image is captured multiple times, similar to the image with a part, while changing the imaging conditions (side illumination, incident illumination, and illumination patterns with combinations of R, G, and B colored light). Note that confirmation of whether or not an image of the part-free image has been captured may be done by visual inspection by an operator. Alternatively, confirmation of whether or not an image of the part-free image has been captured may be done by imaging the tip of the suction nozzle 15 after the suction operation using a parts camera 18 or the like, after the suction operation has been performed and the image of the part-free image has been captured, and confirming whether or not a part has been picked up by the suction nozzle 15.

[0032] When the CPU 42 captures images with and without parts multiple times under different imaging conditions, it performs the above imaging operation multiple times each time a pick-up operation is performed on the same target feeder 30 (S100~S140). As a result, multiple images with and without parts are acquired for each imaging condition. The number of sets can be set as appropriate by the operator. Alternatively, the CPU 42 may perform the above imaging operation only once. In other words, the CPU 42 may acquire one image with and one image without parts for each imaging condition.

[0033] The CPU 42 then extracts multiple types of feature quantities from the image with and without parts for each imaging condition of the target feeder 30 (S150). Feature quantity extraction is performed by setting a predetermined range including the part supply position (cavity 34) as the extraction range for each image (image with parts, image without parts), obtaining the brightness of each pixel within the set extraction range, and calculating multiple types of feature quantities based on the obtained brightness. Examples of multiple types of feature quantities include maximum brightness, minimum brightness, average brightness, contrast, variance, and standard deviation. Contrast can be obtained, for example, by the ratio of maximum brightness to minimum brightness. Variance can be obtained by dividing the sum of the squares of the deviations between the brightness of each pixel and the average brightness by the number of pixels.

[0034] Next, the CPU 42 calculates a feature ratio (feature ratio) for each combination of imaging conditions and feature quantities by dividing the larger of the feature quantities of the image with parts and the feature quantities of the image without parts by the smaller of the two (S160). Subsequently, the CPU 42 selects the combination with the maximum feature ratio among the combinations of imaging conditions and feature quantities as an effective combination for determining the presence or absence of parts (S170). The feature ratio may also be calculated by dividing the smaller of the feature quantities of the image with parts and the feature quantities of the image without parts by the larger of the two. In this case, the determination in S170 only requires selecting the combination with the minimum feature ratio as the effective combination. Then, the CPU 42 registers the imaging conditions and feature quantities of the effective combination as effective imaging conditions and effective feature quantities, respectively, in the storage unit 48 and associates them with the target feeder 30 (S180), and ends the pre-determination processing.

[0035] Figure 6 shows the minimum brightness of R-light side illumination for images with and without parts. Figure 7 shows the contrast of images with and without parts with B-light fully illuminated. Figure 8 shows the maximum brightness of images with and without parts with G-light incident illumination. Figure 9 shows the average brightness of images with and without parts with R-light incident illumination. These figures compare identical feature quantities between multiple images with and without parts, each captured before and after multiple adsorption operations under the same imaging conditions. The feature quantity ratio was calculated, for example, by dividing the minimum value (C0 in the figures) of the feature quantities extracted from multiple images without parts captured under the same imaging conditions by the maximum value (C1 in the figures) of the feature quantities extracted from multiple images with parts captured under the same imaging conditions, as shown in Figures 6 to 9. Furthermore, if there is one image with a component and one image without a component captured under the same imaging conditions, the feature ratio can be calculated by dividing the features extracted from one image without a component by the features extracted from one image with a component captured under the same imaging conditions. In the examples in Figures 6 to 9, the combination of imaging conditions and feature types that yields the largest feature ratio is the combination where the imaging condition type is R-light side emission and the feature type is minimum brightness (see Figure 6). In this case, R-light side emission becomes the effective imaging condition, and minimum brightness becomes the effective feature, and this information is registered in the storage unit 48 as judgment information 49, associated with the target feeder 30. An example of judgment information 49 for each feeder is shown in Figure 10. Judgment information 49 is created for each feeder 30 and, as shown in the figure, is registered in the storage unit 48, associated with identification information (feeder number) that identifies the feeder 30.

[0036] Next, we will explain the component presence / absence determination process. Figure 11 is a flowchart showing an example of the component presence / absence determination process executed by the CPU 42 of the control device 40.

[0037] In the component presence / absence determination process, the CPU 42 first determines whether the effective imaging conditions and effective feature quantities for the target feeder 30 that supplies the target component are already registered in the storage unit 48 (determination information 49) (S200). If the CPU 42 determines that the effective imaging conditions and effective feature quantities are not already registered for the target feeder 30, it terminates the component presence / absence determination process.

[0038] On the other hand, if the CPU 42 determines that effective imaging conditions and effective features have been registered for the target feeder 30, it determines whether or not the suction operation has not yet been performed (S210). If the CPU 42 determines that the suction operation has not yet been performed, it refers to the feeder-specific determination information 49 stored in the memory unit 48 and obtains the effective imaging conditions corresponding to the target feeder 30 (S220). Then, the CPU 42 controls the mark camera 20 to irradiate light from the illumination unit 21 with the obtained effective imaging conditions and capture an image of the tape 32 (cavity 34) of the target feeder 30 (S230). Based on the image of the cavity 34 (the image to be determined) captured in S230, the CPU 42 determines whether or not there are any components in the cavity 34.

[0039] Next, the CPU 42 refers to the feeder-specific determination information 49 to obtain the type of effective feature corresponding to the target feeder 30 (S240), and extracts the type of effective feature obtained in S240 from the captured image of the target to be determined (S250). Subsequently, the CPU 42 refers to the feeder-specific determination information 49 to obtain the effective feature quantities of the image with parts and the image without parts corresponding to the target feeder 30 (S260), and performs a similarity determination to determine whether the effective feature quantities of the image to be determined are similar to those of the effective feature quantities of the image with parts or the effective feature quantities of the image without parts (S270). The similarity determination can be performed, for example, as follows: The CPU 42 calculates the distance between the effective feature quantities of the image to be determined and the effective feature quantities of the image with parts when they are displayed as points in a 2D coordinate system. Similarly, the CPU 42 calculates the distance between the effective feature quantities of the image to be determined and the effective feature quantities of the image without parts when they are displayed as points in a 2D coordinate system. Next, CPU42 compares the lengths of the distances between two points. Then, CPU42 determines that the image to be judged is similar to the image with parts if the distance between the effective features of the image to be judged and the effective features of the image with parts is shorter than the distance between the effective features of the image to be judged and the effective features of the image without parts. On the other hand, CPU42 determines that the image to be judged is similar to the image without parts if the distance between the effective features of the image to be judged and the effective features of the image without parts is shorter than the distance between the effective features of the image to be judged and the effective features of the image with parts.

[0040] If the CPU 42 determines that the image to be judged is similar to an image with a component, it determines that there is a component in the cavity 34 (S290), proceeds to the suction operation (S300), and terminates the component presence / absence determination process. On the other hand, if the CPU 42 determines that the image to be judged is similar to an image without a component, it determines that there is no component in the cavity 34 (S310), outputs an error without proceeding to the suction operation (S320), and terminates the component presence / absence determination process.

[0041] Here, for example, if a transparent tape (transparent embossed tape) is used as tape 32, the way light reflects may differ for each feeder 30 due to differences in production lots and scratches from aging, resulting in variations in how the images appear with and without parts. In this case, depending on the type of feature, the feature quantities obtained for images with and without parts may be similar, making it difficult to determine the presence or absence of parts. In this embodiment, the CPU 42 pre-registers, for each feeder 30, the type of feature quantity that has the highest ratio (feature ratio) between the feature quantities in images with and without parts, as an effective feature quantity for determining the presence or absence of parts. This allows for the extraction of the optimal type of feature quantity from the images on tape 32 for each feeder 30, enabling the determination of the presence or absence of parts from the images on tape 32 with good accuracy.

[0042] Furthermore, when imaging a component whose surface is covered with a colored film (for example, a blue film), depending on the color of the light emitted from the illumination unit 21 (for example, red), the image may appear dark, making it difficult to obtain feature data. For this reason, effective imaging conditions and effective feature data may be selected for each component type, either separately or in addition to each feeder 30.

[0043] Here, we will explain the correspondence between the main elements of the embodiment and the main elements of the present disclosure as described in the claims. Specifically, the CPU 42 of the control device 40 that performs the pre-determination processing of the embodiment corresponds to the selection unit of the present disclosure, the storage unit 48 of the control device 40 corresponds to the storage unit, and the CPU 42 of the control device 40 that performs the component presence / absence determination processing corresponds to the determination unit.

[0044] It goes without saying that this disclosure is not limited in any way to the embodiments described above, and can be implemented in various forms as long as they fall within the technical scope of this disclosure.

[0045] For example, in the embodiment described above, the CPU 42 selects an effective illumination pattern from among several types of illumination patterns as the image acquisition condition for determining the presence or absence of a component. However, the CPU 42 only needs to select an effective imaging condition from among several types of imaging conditions, such as selecting a shutter speed that is effective for the determination from among several different shutter speeds. Alternatively, the CPU 42 may omit the selection of effective imaging conditions for each feeder 30 and apply a common imaging condition to all feeders 30.

[0046] Furthermore, in the embodiment described above, the CPU 42 selects an effective feature (effective feature) from among multiple types of feature quantities based on the ratio of the feature quantities of the image with the part to the feature quantities of the image without the part. However, when the CPU 42 selects an effective feature from among multiple types of feature quantities, for example, maximum brightness, minimum brightness, and average brightness, it may select the feature quantity with the largest difference between the feature quantities of the image with the part and the feature quantities of the image without the part as the effective feature.

[0047] Furthermore, in the embodiment described above, the CPU 42 selects one effective feature from among multiple types of features that is effective for the determination, but it may also select two or more effective features. In this case, the CPU 42 may, for example, select as an effective feature a feature whose ratio of the features of the image with parts to the features of the image without parts is within a predetermined range (for example, the value obtained by dividing the larger of the features of the image with parts and the features of the image without parts by the smaller is greater than or equal to a predetermined value). Alternatively, the CPU 42 may select as an effective feature a feature whose difference between the features of the image with parts and the features of the image without parts is greater than or equal to a predetermined value. When determining the presence or absence of parts using two or more effective features, the CPU 42 may perform a similarity determination of the target image for each of the two or more effective features, comparing the image with parts and the image without parts, and then determine the presence or absence of parts based on the results of each similarity determination.

[0048] As explained above, in the part presence / absence determination method of this disclosure, for each feeder or part type, one or more types of feature quantities effective for determination are selected from among multiple types of feature quantities based on multiple types of feature quantities extracted from images with and without parts. Then, in determining whether or not a target part is present in the cavity of the target feeder, the determination is made using the type of feature quantity selected for the target feeder or the part type of the target part. As a result, regardless of the feeder or part type, the presence or absence of a part in the cavity can be determined from the tape image with good accuracy.

[0049] In the part presence / absence determination method of this disclosure, the type of feature quantity effective for the determination may be selected based on the ratio or difference between the feature quantity of the image with the part and the feature quantity of the image without the part. Alternatively, multiple images with and without parts may be acquired, and the type of feature quantity effective for the determination may be selected based on the ratio or difference between the feature quantities of the acquired multiple images with and without parts and the feature quantities of the two closest points to each other. In this way, a feature quantity effective for the determination can be selected from multiple types of feature quantities for each feeder or part type with a simple process.

[0050] Furthermore, in the part presence / absence determination method of this disclosure, for each feeder or part type, images of the part present and images of the part absent may be captured under multiple types of imaging conditions, multiple types of feature quantities may be extracted from the captured images of the part present and images of the part absent, one or more combinations of imaging conditions and feature quantities that are effective for the determination may be selected from the combinations of multiple types of imaging conditions and multiple types of feature quantities, and the selected combination may be stored in advance as an effective combination. When determining whether or not a target part is present in the cavity of a target feeder that is the target of determination among the multiple feeders, an image of the cavity may be captured under the effective combination of types selected for the target feeder or the part type of the target part, and the effective feature quantities may be extracted from the captured image, and the presence or absence of the target part may be determined based on the extracted effective feature quantities and the corresponding effective feature quantities of the images of the part present and images of the part absent, which have been stored in advance. In this way, by selecting effective imaging conditions in addition to effective feature quantities, the presence or absence of a part in the cavity can be determined from the tape image with even better accuracy.

[0051] Furthermore, the present disclosure described above is not limited to the form of a part presence / absence determination method, but may also be in the form of an image processing system including a control device 40 that performs pre-determination processing and part presence / absence determination processing. [Industrial applicability]

[0052] This disclosure can be used in industries such as the manufacturing of image processing systems. [Explanation of Symbols]

[0053] 10 Component mounting machine, 11 Base, 12 Board transport device, 14 Mounting head, 15 Suction nozzle, 16 Head moving device, 16a X-axis slider, 16b Y-axis slider, 18 Part camera, 20 Mark camera, 21 Illumination unit, 22 Side illumination unit, 22b Blue LED, 22g Green LED, 22r Red LED, 23 Incident illumination unit, 23b Blue LED, 23g Green LED, 23r Red LED, 24 Half mirror, 25 Lens, 26 Imaging unit, 30 Feeder, 32 Tape, 33 Sprocket hole, 34 Cavity, 40 Control device, 42 CPU, 44 ROM, 46 RAM, 48 Storage unit, 49 Judgment information, 50 Management device, 52 Management control unit, 54 Input device, 56 Display, 58 Storage unit.

Claims

1. A method for determining whether or not a component is present in a cavity of a tape, in a feeder that supplies a tape having multiple cavities for housing components, wherein the method determines from an image of the tape whether or not a component is present in the cavity. For each feeder or part type, multiple types of feature quantities are extracted from the image with a part in the cavity and the image without a part in the cavity, and based on the extracted multiple types of feature quantities, two or more types of feature quantities that are effective for the determination are selected from among the multiple types of feature quantities, and the selected types of feature quantities are stored in advance as effective feature quantities. In determining whether or not a target component is present in the cavity of a target feeder among the plurality of feeders, an image of the tape of the target feeder is acquired, effective feature quantities of a selected type for the target feeder or the type of component of the target component are extracted from the acquired image, and it is determined whether or not the target component is present in the cavity of the target feeder based on the extracted effective feature quantities and the corresponding effective feature quantities of the image with the component and the image without the component, which have been stored in advance. A feature whose ratio to the feature of the image with the part is within a predetermined range, or a feature whose difference between the feature of the image with the part and the feature of the image without the part is greater than or equal to a predetermined value, is selected as the type of feature effective for the determination. Method for determining the presence or absence of a part.

2. A method for determining the presence or absence of a part according to claim 1, Multiple images with the component and multiple images without the component are acquired, and the type of feature that is effective for the determination is selected based on the ratio or difference between the two closest points among the acquired feature quantities of the multiple images with the component and the feature quantities of the multiple images without the component. Method for determining the presence or absence of a part.

3. A method for determining the presence or absence of a part according to claim 1 or 2, For each feeder or part type, images with and without the part are captured under multiple imaging conditions, multiple types of feature quantities are extracted from the captured images with and without the part, one or more combinations of imaging conditions and feature quantities that are effective for the determination are selected from the multiple combinations of imaging conditions and feature quantities, and the selected combinations are stored in advance as effective combinations. In determining whether or not a target component is present in the cavity of a target feeder among the plurality of feeders, an image of the cavity is captured using an effective combination of types selected for the target feeder or the type of the target component, and effective feature quantities are extracted from the captured image. Based on the extracted effective feature quantities and the corresponding effective feature quantities of the image with the component and the image without the component stored in advance, it is determined whether or not the target component is present in the cavity of the target feeder. Method for determining the presence or absence of a part.

4. An image processing system for determining whether or not there are components in a cavity in a tape feeder that supplies a tape having multiple cavities for housing components, wherein A selection unit extracts multiple types of feature quantities from images showing parts in the cavity and images showing parts not in the cavity, respectively, for each feeder or part type, and selects two or more feature quantities that are effective for the determination from among the extracted multiple types of feature quantities. A storage unit that stores the type of feature quantity selected by the selection unit as an effective feature quantity, In determining whether or not a target component is present in the cavity of a target feeder among the plurality of feeders, the determination unit acquires an image of the tape of the target feeder, extracts effective feature quantities of the type selected by the selection unit for the target feeder or the type of component of the target component from the acquired image, and determines whether or not the target component is present in the cavity of the target feeder based on the extracted effective feature quantities and the corresponding effective feature quantities of the image with the component and the image without the component stored in the storage unit, respectively. Equipped with, The selection unit is an image processing system that selects as the type of feature effective for the determination a feature whose ratio to the feature of the image without the part is within a predetermined range, or a feature whose difference to the feature of the image without the part is greater than or equal to a predetermined value.

Images