Manufacturing methods for inspection devices, cutting devices, and semiconductor components

Abstract

This invention provides an inspection apparatus that appropriately sets defect mark regions within an image of an object to be inspected, thereby improving the accuracy of defect mark inspection. The inspection apparatus includes a control unit. The control unit pre-sets a defect mark region (area where defect marks are captured if the object contains defect marks) within a pre-set image containing the object to be inspected, and inspects this defect mark region within the inspection image to determine whether the object to be inspected contains defect marks. Pre-setting the defect mark region includes acquiring pre-set images containing both good and defective objects (objects without defect marks), comparing the brightness of the good object region (containing good objects) with the brightness of the defective object region (containing defective objects) within the pre-set image, and thereby determining the defect mark region.

Description

Technical Field

[0001] The invention relates to an inspection device, a cutting device and a method for manufacturing a semiconductor component. Prior Art

[0002] In production lines for products such as semiconductor parts, products are sometimes labeled with bad marks indicating that the products are defective. In this case, a subsequent inspection device performs a bad mark inspection multiple times to check whether the products are labeled with bad marks. Patent document 1 discloses a bad mark detection method based on image processing. [Prior art literature] [Patent Document]

[0003] [Patent Document 1] Japanese Patent Application Publication No. 2013-38285 Summary of the invention

[0004] [Problems to be solved by the invention]

[0005] In order to accurately detect bad marks, it is desirable to set the area where the bad marks are marked if the product contains bad marks (bad mark area) as accurately as possible in the image of the product. However, since the position and shape of the bad marks marked vary depending on the product, the position of the bad mark area in the image cannot be determined uniformly in many cases. In addition, the position of the bad mark area in the image will also change depending on the installation position or angle of the camera. Therefore, in many cases, the bad mark area has to be set manually. As a result, there are many problems such as deviations in the settings performed by the operator, making it difficult to ensure the accuracy of the bad mark inspection.

[0006] The object of the present invention is to provide an inspection device, a cutting device and a method for manufacturing a semiconductor component, wherein the inspection device appropriately sets a bad mark area in an image of an inspection object, thereby improving the accuracy of bad mark inspection. [Methods to solve the problem]

[0007] The inspection device according to one aspect of the present invention includes a control unit. The control unit pre-sets, in a setting image captured of the inspection object, an area where a defective mark is captured when the inspection object contains a defective mark, i.e., a defective mark area, and inspects the defective mark area in the inspection image, thereby inspecting whether the inspection object captured in the inspection image contains a defective mark. The case of pre-setting the defective mark area includes: obtaining a setting image in which an inspection object that does not contain a defective mark, i.e., a good product, and an inspection object that contains a defective mark, i.e., a defective product, are captured, and comparing the brightness of the good product area captured in the setting image with the brightness of the defective product area captured in the setting image, thereby determining the defective mark area.

[0008] According to another aspect of the present invention, a cutting device comprises: a cutting mechanism for cutting a substrate including a plurality of semiconductor components and singulating the semiconductor components; and the inspection device for inspecting the singulated semiconductor components as inspection objects.

[0009] According to another aspect of the present invention, a method for manufacturing semiconductor parts includes the following steps: cutting a substrate including a plurality of semiconductor parts by a cutting mechanism and singulating the semiconductor parts; and inspecting the singulated semiconductor parts as inspection objects using the inspection device. [Effects of the Invention]

[0010] According to the present invention, the defective mark area is appropriately set in the image of the inspection object, thereby improving the accuracy of the defective mark inspection. Simple diagram description

[0011] FIG. 1 is a plan view schematically showing a cutting device according to an embodiment. FIG. 2 is a side cross-sectional view schematically showing an imaging environment by a second optical inspection camera according to an embodiment. FIG. 3 is a diagram schematically showing a hardware configuration of a computer according to an embodiment. FIG. 4 is a diagram showing an example of an image of a semiconductor component captured by a second optical inspection camera according to an embodiment. FIG. 5 is a flowchart showing the flow of a method for setting parameters for defective mark inspection according to one embodiment. FIG. 6 is a diagram showing an example of divided regions set in one component region according to one embodiment. FIG. 7A is a diagram showing a comparison result of brightness between a good product region and a defective product region according to an embodiment. FIG. 7B is a diagram showing a comparison result of brightness between a good product region according to one embodiment and another good product region. FIG. 7C is a diagram showing a comparison result of brightness between a good product region according to one embodiment and another good product region. FIG. 8 is a table showing calculation results for setting the third threshold value according to one embodiment. FIG. 9 is a graph showing the relationship between parameters and area differences according to one embodiment. Implementation

[0012] Hereinafter, an embodiment of one aspect of the present invention (hereinafter, also referred to as "this embodiment") is described in detail using drawings. In addition, the same or corresponding parts in the drawings are marked with the same symbols and their descriptions are not repeated. In addition, in each drawing, objects are appropriately omitted or exaggerated for easy understanding.

[0013] [1.Structure] <1-1. Overall structure of the cutting device> Fig. 1 is a schematic plan view of a cutting device 1 according to the present embodiment. The cutting device 1 is configured to separate a package substrate (cutting object) into a plurality of semiconductor components (package components) by cutting the package substrate. In the package substrate, a substrate or a lead frame on which a semiconductor chip is mounted is sealed with resin.

[0014] As an example of a packaging substrate, there are: a ball grid array (BGA) packaging substrate, a land grid array (LGA) packaging substrate, a chip size package (CSP) packaging substrate, a light emitting diode (LED) packaging substrate, and a quad flat no-leaded packaging substrate.

[0015] In addition, the cutting device 1 is configured to inspect each of the plurality of semiconductor components after singulation. In the cutting device 1, an image of each semiconductor component is captured, and each semiconductor component is inspected based on the image. Inspection data is generated through the inspection, and each semiconductor component is classified as a "good product" or a "defective product".

[0016] In the example described above, a package substrate P1 is used as a cutting object, and the package substrate P1 is singulated into a plurality of semiconductor components S1 by a cutting device 1. Hereinafter, the surface of the package substrate P1 sealed with resin is referred to as a molding surface, and the surface opposite to the molding surface is referred to as a ball / lead surface.

[0017] As shown in FIG1 , the cutting device 1 includes a cutting module (cutting mechanism) A1 and an inspection / storage module (inspection device) B1 as components. The cutting module A1 is configured to manufacture a plurality of semiconductor components S1 by cutting a package substrate P1. The inspection / storage module B1 is configured to inspect each of the manufactured plurality of semiconductor components S1 as an inspection object, and then store the semiconductor components S1 in a tray. In the cutting device 1, each component can be detached from other components and can be replaced.

[0018] The cutting module A1 mainly includes a substrate supply unit 3 , a positioning unit 4 , a cutting table 5 , a spindle unit 6 , and a conveying unit 7 .

[0019] The substrate supply unit 3 pushes out the package substrates P1 one by one from the magazine M1 accommodating a plurality of package substrates P1, thereby supplying the package substrates P1 one by one to the positioning unit 4. At this time, the package substrates P1 are arranged with the ball / lead surface facing upward.

[0020] The positioning section 4 positions the package substrate P1 pushed out from the substrate supply section 3 on the rail section 4 a . Thereafter, the positioning section 4 conveys the positioned package substrate P1 to the cutting table 5 .

[0021] The cutting table 5 holds the package substrate P1 to be cut. In the example, a cutting device 1 having a double cutting table structure with two cutting tables 5 is illustrated. The cutting table 5 includes a holding member 5a, a rotating mechanism 5b, and a moving mechanism 5c. The holding member 5a holds the package substrate P1 by adsorbing the package substrate P1 conveyed by the positioning unit 4 from below. The rotating mechanism 5b can rotate the holding member 5a in the θ1 direction of FIG. 1 (i.e., rotate in the XY horizontal plane of FIG. 1). The moving mechanism 5c can move the holding member 5a along the Y axis of the figure.

[0022] The spindle unit 6 separates the package substrate P1 into a plurality of semiconductor components S1 by cutting the package substrate P1. In the above example, a cutting device 1 having a dual spindle structure with two spindle units 6 is illustrated. The spindle unit 6 can move along the X-axis and the Z-axis of the figure. Furthermore, the cutting device 1 can also adopt a single spindle structure having one spindle unit 6.

[0023] The main shaft portion 6 includes a blade 6a and a rotating shaft 6c. The blade 6a cuts the package substrate P1 by rotating at a high speed, thereby singulating the package substrate P1 into a plurality of semiconductor components S1. The blade 6a is mounted on the rotating shaft 6c in a state of being clamped by a first flange and a second flange (not shown). The first flange and the second flange are fixed to the rotating shaft 6c by a fastening member (not shown) such as a nut. The first flange is also called an inner flange, and the second flange is also called an outer flange.

[0024] The main shaft 6 is provided with a cutting water nozzle, a cooling water nozzle, and a cleaning water nozzle (none of which is shown). The cutting water nozzle sprays cutting water toward the blades 6a rotating at high speed. The cooling water nozzle sprays cooling water. The cleaning water nozzle sprays cleaning water for cleaning the cutting chips.

[0025] Referring again to FIG. 1 , after the cutting workbench 5 absorbs the package substrate P1, the first position confirmation camera 5d photographs the package substrate P1 to confirm the position of the package substrate P1. The confirmation using the first position confirmation camera 5d is, for example, confirmation of the position of a mark provided on the package substrate P1. The mark is, for example, a mark for determining the cutting position of the package substrate P1.

[0026] Thereafter, the cutting table 5 moves toward the main shaft portion 6 along the Y axis of the figure. After the cutting table 5 moves to the bottom of the main shaft portion 6, it is aligned, and then the package substrate P1 is cut by moving the cutting table 5 and the main shaft portion 6 relative to each other. Each time the package substrate P1 is cut by the blade 6a of the main shaft portion 6, the package substrate P1 is photographed and confirmed by the second position confirmation camera 6b provided on the main shaft portion 6. The confirmation using the second position confirmation camera 6b is, for example, confirmation of the cut position and the cut width of the package substrate P1.

[0027] After the cutting of the package substrate P1 is completed, the cutting table 5 moves along the Y-axis of the figure in a direction away from the main shaft portion 6 while adsorbing the plurality of singulated semiconductor components S1. During the movement, the upper surface (ball / lead surface) of the semiconductor component S1 is cleaned and dried by the first cleaner 5e.

[0028] The conveyor 7 sucks the semiconductor component S1 held by the cutting stage 5 from above and conveys the semiconductor component S1 to the inspection stage 11 of the inspection / storage module B1. During the conveying process, the lower surface (molding surface) of the semiconductor component S1 is cleaned and dried by the second cleaner 7a.

[0029] The inspection / storage module B1 mainly includes an inspection workbench 11, a first optical inspection camera 12, a second optical inspection camera 13, an illumination unit 16, an illumination unit 17, a configuration unit 14, and an extraction unit 15. Furthermore, the first optical inspection camera 12 may also be disposed in the cutting module A1.

[0030] In order to perform optical inspection of the semiconductor component S1, the inspection table 11 holds the semiconductor component S1. The inspection table 11 can move along the X-axis of the figure. In addition, the inspection table 11 can be turned upside down. A holding member is provided on the inspection table 11, and the holding member holds the semiconductor component S1 by adsorbing the semiconductor component S1. In addition, in the inspection table 11, the surface for holding the semiconductor component S1 includes a dark-colored rubber such as black in the example described above. Furthermore, the color of the rubber does not need to be a dark color, and it can also be a bright color such as white.

[0031] The first optical inspection camera 12 and the second optical inspection camera 13 respectively photograph the molded surface and the ball / lead surface of the semiconductor component S1. Various inspections of the semiconductor component S1 are performed based on the images (image data) generated by the first optical inspection camera 12 and the second optical inspection camera 13. Each of the first optical inspection camera 12 and the second optical inspection camera 13 is arranged near the inspection workbench 11 to photograph the upper side. In the example described above, the image generated by each of the first optical inspection camera 12 and the second optical inspection camera 13 is a gray scale (256 gray levels) image, but it is not limited to this, and for example, it can also be a color image. Furthermore, after the package substrate P1 is cut by the spindle part 6, the plurality of semiconductor components S1 after singulation maintain the arrangement of the package substrate P1 before cutting until they are photographed by the first optical inspection camera 12 and the second optical inspection camera 13, and then arranged in the arrangement part 14 as described later. That is, during this period, the plurality of semiconductor components S1 after singulation maintain the state of being arranged vertically and horizontally. Therefore, the semiconductor components S1 are captured in the images captured by the first optical inspection camera 12 and the second optical inspection camera 13 in a state of being arranged vertically and horizontally.

[0032] The first optical inspection camera 12 photographs the molded surface of the semiconductor component S1 transported to the inspection table 11 by the transport unit 7. Thereafter, the transport unit 7 places the semiconductor component S1 on the holding member of the inspection table 11. The inspection table 11 is turned upside down after adsorbing the semiconductor component S1. The inspection table 11 moves above the second optical inspection camera 13, and the second optical inspection camera 13 photographs the ball / lead surface of the semiconductor component S1.

[0033] An illumination unit 16 is provided above the first optical inspection camera 12, and an illumination unit 17 is provided above the second optical inspection camera 13. Each of the illumination units 16 and 17 includes, for example, so-called coaxial illumination and / or dome-shaped illumination. The illumination unit 16 is configured to irradiate light to the semiconductor component S1 on the conveyor 7 during inspection by the first optical inspection camera 12. The illumination unit 17 is configured to irradiate light to the semiconductor component S1 on the inspection table 11 during inspection by the second optical inspection camera 13. The combination of the first optical inspection camera 12 and the illumination unit 16 and the combination of the second optical inspection camera 13 and the illumination unit 17 have, for example, the same structure, and therefore the structure of the latter combination will be described representatively below.

[0034] FIG. 2 is a side cross-sectional view schematically showing a photographing environment by the second optical inspection camera 13. As shown in FIG. 2, in the inspection by the second optical inspection camera 13, light emitted by the lighting unit 17 is irradiated to the semiconductor component S1. In the example, the lighting unit 17 includes a dome-shaped lighting including a dome 17a and a plurality of LEDs 17b arranged on the inner surface of the dome 17a. In a state where light is irradiated to the semiconductor component S1, an image of the semiconductor component S1 is generated by the second optical inspection camera 13. Based on the image, the semiconductor component S1 is inspected.

[0035] 1 again, the semiconductor component S1 after inspection is arranged in the arrangement section 14. The arrangement section 14 is movable along the Y axis of the figure. The inspection table 11 arranges the semiconductor component S1 after inspection in the arrangement section 14.

[0036] The extraction unit 15 transfers the semiconductor components S1 arranged in the arrangement unit 14 to the tray. Based on the inspection results using the first optical inspection camera 12 and the second optical inspection camera 13, the semiconductor components S1 are classified as "good" or "defective". Based on the results of the classification, the extraction unit 15 transfers each semiconductor component S1 to the good tray 15a or the defective tray 15b. That is, good components are stored in the good tray 15a, and defective components are stored in the defective tray 15b. When each of the good tray 15a and the defective tray 15b is filled with semiconductor components S1, it is replaced with a new tray.

[0037] The cutting device 1 further includes a computer 50 and a monitor 20. The monitor 20 is configured to display an image. The monitor 20 includes, for example, a display device such as a liquid crystal monitor or an organic electroluminescence (EL) monitor. Furthermore, in the example of FIG. 1 , the computer 50 and the monitor 20 are disposed in the inspection / storage module B1, but may also be disposed in the cutting module A1.

[0038] The computer 50 controls the operation of each part of the cutting module A1 and the inspection / storage module B1, for example. The computer 50 controls the operation of the substrate supply unit 3, the positioning unit 4, the cutting table 5, the spindle unit 6, the conveying unit 7, the inspection table 11, the first position confirmation camera 5d, the second position confirmation camera 6b, the first optical inspection camera 12, the second optical inspection camera 13, the lighting unit 16, the lighting unit 17, the arrangement unit 14, the extraction unit 15, and the monitor 20, for example.

[0039] In addition, the computer 50 performs various inspections of the semiconductor component S1 based on, for example, image data generated by the first optical inspection camera 12 and the second optical inspection camera 13. Next, the computer 50 will be described in detail.

[0040] <1-2. Computer hardware structure> Fig. 3 is a diagram schematically showing the hardware structure of the computer 50. As shown in Fig. 3, the computer 50 includes a control unit 70, an input / output interface (I / F) 90, a receiving unit 95, and a storage unit 80, and each structure is electrically connected via a bus.

[0041] The control unit 70 includes a central processing unit (CPU) 72, a random access memory (RAM) 74, and a read-only memory (ROM) 76. The control unit 70 is configured to control each component in the computer 50 according to information processing, thereby controlling each component in the cutting device 1 except the computer 50.

[0042] The input / output I / F 90 is configured to communicate with each component included in the cutting device 1 via a signal line. The input / output I / F 90 is used to send data from the computer 50 to each component in the cutting device 1, and receive data sent from each component in the cutting device 1 to the computer 50. The receiving unit 95 is configured to receive instructions from the user. The receiving unit 95 includes, for example, a part or all of a touch panel, a keyboard, a mouse, and a microphone.

[0043] The storage unit 80 is, for example, an auxiliary storage device such as a hard disk drive or a solid state drive. The storage unit 80 is configured to store, for example, a control program 81. The various actions of the cutting device 1 are realized by the control unit 70 executing the control program 81. When the control unit 70 executes the control program 81, the control program 81 is expanded in the RAM 74. Then, the control unit 70 controls each component by interpreting and executing the control program 81 expanded in the RAM 74 through the CPU 72.

[0044] [2. Method for manufacturing semiconductor components] A method for manufacturing a semiconductor component S1 using the cutting device 1 is described. First, a package substrate P1 is prepared and stored in a material box M1 included in a cutting module A1 of the cutting device 1. In the example, the package substrate P1 is formed by arranging various components such as semiconductor components in a certain repeated pattern on a substrate. Thereby, a package substrate P1 forms a structure in which a plurality of semiconductor components S1 of the same type are connected in a state of being arranged vertically and horizontally. For each semiconductor component S1 included in a package substrate P1, a defective mark is marked as needed before being stored in the material box M1. The defective mark is a mark that is marked on the semiconductor component S1 when it is judged that there are some defects (defective products) in the semiconductor component S1 in a step upstream of the cutting device 1 in the production line. The defective mark is marked on the surface of the semiconductor component S1 in a state that can be visually confirmed from the appearance, for example, by laser cutting, printing, digging, etc.

[0045] The package substrate P1 supplied to the material box M1 is conveyed to the main shaft part 6 in the above-mentioned method and cut by the main shaft part 6. Here, the package substrate P1 is cut vertically and horizontally along the boundary lines of the plurality of semiconductor components S1 included therein. As a result, a plurality of semiconductor components S1 after singulation can be obtained. The plurality of semiconductor components S1 after singulation are conveyed to the inspection / storage module B1 while maintaining the positional relationship before cutting, and various inspections are performed. The various inspections described here include appearance inspections based on images generated by the first optical inspection camera 12 and the second optical inspection camera 13. The appearance inspection based on the image generated by the second optical inspection camera 13 includes a bad mark inspection, which is to inspect whether each semiconductor component S1 captured in the image contains a bad mark. These inspections are performed by the control unit 70 of the computer 50.

[0046] The control unit 70 classifies each semiconductor component S1 as a good product or a defective product through various inspections including a defective mark inspection. The control unit 70 controls the operation of the extraction unit 15 to convey the semiconductor component S1 classified as a good product to the good product tray 15a, and convey the semiconductor component S1 classified as a defective product to the defective product tray 15b. In this way, each semiconductor component S1 is sorted to the good product tray 15a or the defective product tray 15b. In the above step, the semiconductor component S1 judged to contain a defective mark by the defective mark inspection is classified as a defective product and sorted to the defective product tray 15b. On the other hand, the semiconductor component S1 judged not to contain a defective mark is sorted to the good product tray 15a as long as it is not judged as a defective product in other inspections. In the above manner, a plurality of semiconductor components S1 separated from the package substrate P1 and distinguished as good products or defective products can be obtained. In this way, the semiconductor component S1 is manufactured by the cutting device 1.

[0047] [3. Bad mark inspection] Next, the details of the bad mark inspection are described. The following describes the preparation steps performed before the bad mark inspection and the process of the bad mark inspection performed after the preparation steps. The preparation step is usually performed when the package substrate P1 cut by the cutting device 1 is a type that is cut for the first time. The information of the package substrate P1 of the type that has undergone the preparation step once can be saved in the storage unit 80, so that it can be used by calling out from the storage unit 80 when the package substrate P1 of the type is cut next time.

[0048] <3-1. Flow of preparation steps> Before starting the manufacturing step of the semiconductor component S1, a preparation step is performed. In the preparation step, various parameters for operating the cutting device 1 are set. As described above, the manufacturing step of the semiconductor component S1 includes a defective mark inspection. In the preparation step, various parameters for steps other than the defective mark inspection are set, but various parameters for the defective mark inspection are also set. The following describes a method for setting parameters for the defective mark inspection.

[0049] FIG4 shows an example of an image IM1 of a semiconductor component S1 captured by the second optical inspection camera 13. In the example, the image IM1 is obtained by capturing the ball / lead surface side of the semiconductor component S1. A plurality of semiconductor components S1 are captured in the image IM1. In the example, the entirety of four semiconductor components S1 arranged in a 2×2 matrix is captured. The area corresponding to each semiconductor component S1 included in the image IM1 is referred to as a component area. Four component areas T1 to T4 are captured in the image IM1. A plurality of cutting lines formed when the package substrate P1 is singulated into a plurality of semiconductor components S1 are also captured in the image IM1. In the image IM1, the cutting lines appear as an image of the surface on which the inspection table 11 holds the semiconductor component S1. The component areas T1 to T4 are adjacent to each other with such cutting lines as the boundary line BL.

[0050] In the example of FIG4 , only one semiconductor component S1 corresponding to the component area T3 among the four semiconductor components S1 corresponding to the component areas T1 to T4 contains a bad mark BM. In the image IM1, the region where the bad mark BM is captured when the semiconductor component S1 contains the bad mark BM is called the bad mark region region of interest (ROI). The bad mark region ROI in the component area T3 contains the image of the bad mark BM, but the bad mark region ROI in the component areas T1, T2, and T4 does not contain the image of the bad mark BM.

[0051] In order to implement the bad mark inspection with good accuracy, it is desirable to set the bad mark area ROI as accurately as possible within the field of view of the image IM1 of the semiconductor component S1. However, since the position or shape of the bad mark BM marked varies depending on the product, the position of the bad mark area ROI in the image IM1 cannot be determined uniformly in most cases. In addition, due to slight jitters of the installation position or angle of the second optical inspection camera 13, the position of the bad mark area ROI in the image IM1 will also change slightly. Therefore, in the past, although the bad mark area ROI was manually set, there was a problem that there were many deviations in the settings performed by the operator. In addition, the threshold value used to determine whether the bad mark BM is included in the bad mark area ROI was also manually set in the past, but there was also a problem that there were many deviations in the settings performed by the operator. When the bad mark area ROI is manually set in a manner of surrounding the bad mark BM, the judgment of whether it is appropriately surrounded depends on personal subjectivity. In the case of adjusting the threshold value, personal subjectivity is also added. For example, depending on the individual, the bad mark region ROI is set too wide, and the image of the bad mark BM in the bad mark region ROI becomes relatively small. If so, the average brightness in the bad mark region ROI changes, and the bad mark BM may not be detected at the set threshold value. If the bad mark region ROI and the threshold value are not set appropriately as in the above example, the accuracy of the bad mark inspection will be reduced. In addition, if the bad mark region ROI and the threshold value are to be set manually, it will take a long time.

[0052] In view of the above problems, in this embodiment, the bad mark region ROI and the threshold value are automatically set. FIG5 shows the flow of the method for setting the bad mark region ROI and the threshold value (the third threshold value Th3 described later) as parameters for bad mark inspection. The process shown in FIG5 is executed by the control unit 70. The process shown in FIG5 is described in detail below.

[0053] First, in step S100, the control unit 70 cuts the package substrate P1 by the spindle unit 6, similarly to the actual manufacturing of the semiconductor component S1, and then causes the second optical inspection camera 13 to photograph the plurality of semiconductor components S1 after singulation. In the example, in one image IM1 photographed by the second optical inspection camera 13, only a portion of the plurality of semiconductor components S1 included in one package substrate P1 is photographed according to the angle of view of the second optical inspection camera 13. The control unit 70 photographs the arrangement of the plurality of semiconductor components S1 from one package substrate P1 multiple times, thereby acquiring a plurality of images IM1. The image of each semiconductor component S1 from one package substrate P1 is captured in at least one of the plurality of images IM1.

[0054] In the next step S101, the control unit 70 obtains a setting image from the multiple images IM1 obtained in step S100. There may be multiple setting images, but in the description here, for convenience, one setting image is set. Specifically, the control unit 70 causes the monitor 20 to display the multiple images IM1 in sequence according to the operator's operation on the receiving unit 95. The operator operates the receiving unit 95 while checking the monitor 20 to select one image IM1 from the multiple images IM1. The condition of the image IM1 selected here (hereinafter referred to as the selection condition) is that at least one semiconductor component S1 (good product) that does not include a defective mark BM and one semiconductor component S1 (defective product) that includes a defective mark BM are each captured. The control unit 70 obtains the image IM1 that satisfies the selection condition selected by the operator as the setting image. Hereinafter, the setting image is also marked with the symbol IM1.

[0055] In the next step S102, the control unit 70 extracts a plurality of part regions T1 to T4 from the setting image IM1. Furthermore, according to the selection condition, at least one of the part regions T1 to T4 extracted here becomes a good product region where good products are captured, and at least one of the other becomes a defective product region where defective products are captured. In the description here, the image of FIG. 4 is selected as the setting image IM1. In the example of FIG. 4, T1, T2, and T4 represent good product regions, and T3 represents a defective product region. The part regions T1 to T4 can be detected by any method through image processing. For example, the control unit 70 can detect the boundary line BL and cut the part regions T1 to T4 along the boundary line BL.

[0056] In the next step S103, the control unit 70 divides each of the plurality of part areas T1 to T4 including at least one good product area and at least one defective product area into n divided areas D1 to Dn (n is an integer greater than or equal to 2). FIG. 6 shows the divided areas D1 to Dn set in the part area T1. Furthermore, in other part areas T2 to T4, the divided areas D1 to Dn are also set by the same method. In the example, the divided areas D1 to Dn are set in a matrix shape (10×10 in FIG. 6). In addition, in the example, the divided areas D1 to Dn are set evenly without gaps or overlaps in the part area. However, the divided areas D1 to Dn may be set with gaps between them, may overlap each other, or may be set unevenly.

[0057] In the next step S104, the control unit 70 calculates the brightness of each segmented area D1 to Dn in each part area T1 to T4. Each segmented area D1 to Dn includes a plurality of pixels. Therefore, the brightness of a segmented area is a representative value representing the brightness of the plurality of pixels included in the segmented area. In the above example, the average brightness of the plurality of pixels is set as the representative value. However, the representative value may also be set to, for example, the median value or the mode of the brightness of the plurality of pixels.

[0058] In the next step S105, the control unit 70 compares the brightness among the plurality of part areas T1 to T4 including at least one good product area and at least one defective product area. In the example, the brightness comparison is to calculate the difference in brightness (brightness difference). In addition, in the example, one part area to be a reference is selected from the plurality of part areas T1 to T4, and the brightness is compared in a cycle between the selected part area and the remaining three part areas. The one part area to be a reference can be automatically selected by any method such as selection based on the brightness of each part area or random selection, but can also be predetermined. In the example, the part area T1 is selected as the one part area to be a reference, so the brightness is compared for the combination of the three part areas (T1, T2), (T1, T3), and (T1, T4). Alternatively, the brightness can be compared for all the combinations of the plurality of part areas T1 to T4 without selecting one part area to be a reference. According to the above method, the brightness of the good product area must be compared with the brightness of the defective area for at least one group.

[0059] In the example, the brightness is compared for each segmented area D1 to Dn between the part area T1 to the part area T4. Specifically, the difference (brightness difference) between the brightness of the segmented area D1 of one of the two part areas to be compared and the brightness of the segmented area D1 of the other part area is calculated. The same process is repeated for the segmented areas D2 to the segmented areas Dn. Thus, for all the segmented areas D1 to Dn, the difference (brightness difference) between the brightness of the segmented area of one of the part areas and the brightness of the corresponding segmented area in the other part area is calculated. According to the above method, for each segmented area D1 to Dn, the brightness is compared between the good product area and the defective product area. FIG. 7A shows the comparison result of the brightness between the good product area T1 and the defective product area T3. In FIG. 7A, the numerical values in each segmented area D1 to Dn constituting the part area represent the brightness difference calculated for the segmented area.

[0060] In the next step S106, the control unit 70 selects a segmented area where the brightness difference (absolute value) between the two part areas compared in step S105 is greater than the first threshold value Th1. In the example, similarly to step S105, the same process is repeated for the combination of three part areas (T1, T2), (T1, T3), and (T1, T4). Thus, the segmented area where the brightness difference is greater than the first threshold value Th1 must be selected for at least one combination of a good area and a defective area. In FIG7A, the segmented area painted dark gray is the segmented area where the brightness difference between the good area T1 and the defective area T3 is greater than the first threshold value Th1. Although the numerical value of the brightness difference is omitted, FIG7B shows the comparison result of the brightness between the good area T1 and the good area T2, and FIG7C shows the comparison result of the brightness between the good area T1 and the good area T4. In FIG. 7B and FIG. 7C , similarly to FIG. 7A , the segmented regions painted dark gray are segmented regions where the brightness difference between the two component regions is greater than or equal to the first threshold value Th1 .

[0061] The first threshold value Th1 can be set appropriately. For example, the first threshold value Th1 can be a fixed value or can be set automatically. In the case of automatic setting, for example, the first threshold value Th1 can be calculated based on the brightness of the two target part regions. In the example, the first threshold value Th1 is automatically set to (the maximum value of the brightness difference) / 2.

[0062] In the next step S107, the control unit 70 further selects a segmented area in which the brightness difference between the component areas is greater than the second threshold value Th2 among the segmented areas adjacent to the segmented area selected in step S106. In the example, the same process is repeated for the combination of three component areas (T1, T2), (T1, T3), and (T1, T4) as in step S105 and step S106. In FIGS. 7A to 7C, the segmented area painted in light gray is the segmented area selected in step S107.

[0063] The second threshold value Th2 can also be appropriately set. In the example described above, the second threshold value Th2 is a fixed value. However, the second threshold value Th2 can also be automatically set. In the case of automatic setting, for example, the second threshold value Th2 can be calculated based on the brightness of the two part areas as the object, and can also be calculated based on the first threshold value Th1. The second threshold value Th2 is preferably smaller than the first threshold value Th1. Therefore, for example, the second threshold value Th2 can also be set to a value obtained by multiplying the first threshold value Th1 by a coefficient greater than 0 and less than 1. In the case where the second threshold value Th2 is set to a fixed value and the first threshold value Th1 is automatically set, it is preferably to predict the value that the first threshold value Th1 can take and pre-set the second threshold value Th2 smaller than it.

[0064] Here, if step S106 is reviewed with reference to FIG. 7A , it can be considered that the segmented area where the brightness difference (absolute value) between the so-called good product area T1 and the defective product area T3 is greater than the first threshold value Th1 (i.e., becomes larger) is the area where the bad mark BM mainly exists. It can be considered that the segmented area selected in step S106, that is, the area around which the bad mark BM is likely to exist, is also likely to exist. Therefore, for the purpose of leak-proof detection of the segmented area that is not detected in step S106 but is likely to exist, the step S107 is performed. The case where the second threshold value Th2 is set to a value smaller than the first threshold value Th1 means that the detection sensitivity is increased in step S107 compared with the selection in step S106 according to the above purpose. The area selected in step S107 is an area with a relatively large brightness difference, and is also considered to be an area where the bad mark BM mainly exists.

[0065] In the next step S108, the control unit 70 determines the bad mark region ROI in a manner including the segmented regions selected in step S106 and step S107. In the example, first, for the three combinations of (T1, T2), (T1, T3), and (T1, T4), one or more regions (hereinafter referred to as combined regions) C formed by connecting the selected segmented regions are detected. Then, the maximum combined region Cmax of the combined region C with the largest area among the detected one or more combined regions C is detected, and the bad mark region ROI is determined in a manner including the maximum combined region Cmax. In the example of FIG. 7A to FIG. 7C, a total of 10 combined regions C are detected. Then, the region with the largest area among them is detected as the maximum combined region Cmax. The maximum combined region Cmax is considered to be an area where the bad mark BM exists. Therefore, the maximum combined region Cmax is detected based on the comparison result (FIG. 7A) between the good product region T1 not including the bad mark BM and the bad product region T3 including the bad mark BM. Due to errors, a combined region C other than the maximum combined region Cmax may be detected.

[0066] When detecting the combined area C, the so-called segmented areas "connected" may mean connected up and down, left and right, or connected up and down, left and right, and obliquely connected. FIG. 7A to FIG. 7C are examples of the case where the segmented areas are connected up and down, left and right, and obliquely connected.

[0067] In the example, as shown in FIG. 7A , the bad mark region ROI is determined as the smallest rectangular region including the maximum bonding region Cmax. However, the example is not limited thereto, and the bad mark region ROI may be, for example, the maximum bonding region Cmax itself, or a smaller region surrounding the maximum bonding region Cmax. In the example, the information for determining the relative position of the bad mark region ROI in the part region is stored in the storage unit 80 as the information for determining the position of the bad mark region ROI.

[0068] By the above steps S100 to S108, the bad mark area ROI is pre-set before the bad mark inspection. The next steps S109 to S111 are the process of pre-setting the third threshold value Th3. The third threshold value Th3 is a parameter for inspecting the bad mark area ROI in the inspection image. The so-called inspection image is an image that is the object of the bad mark inspection, which is photographed by the second optical inspection camera 13 under the same conditions as the setting image IM1. In the example, the third threshold value Th3 is calculated based on the brightness of the bad mark area ROI in the setting image IM1.

[0069] In step S109, the control unit 70 selects a group of a good product area and a bad product area from the plurality of component areas T1 to T4. The present invention is not limited thereto, and the selection is performed by, for example, selecting two component areas T1 and T3 that form the maximum joint area Cmax.

[0070] In the next step S110, the control unit 70 determines an appropriate parameter L as the third threshold value Th3 while changing the parameter L. In the example, the range of change of the parameter L is the maximum range of brightness that the pixel can take (0 to 255). However, the range of change of the parameter L is not limited to this. For example, the range of change of the parameter L is limited to the predicted value of the appropriate value for the third threshold value Th3, for example, it can also be set to a range narrower than the maximum range such as 30 to 230. Furthermore, the scale (increment) of the change of the parameter L can be 1, or it can be a value greater than 2.

[0071] Specifically, the control unit 70 calculates the area of the portion whose brightness exceeds the parameter L in the defective mark region ROI included in one of the part regions (good product region) selected in step S109 while changing the parameter L within the range of variation. Similarly, the control unit 70 calculates the area of the portion whose brightness exceeds the parameter L in the defective mark region ROI included in another part region (defective product region) selected in step S109 while changing the parameter L within the range of variation. For example, the area can be counted as the number of pixels. In addition, the control unit 70 calculates the difference between the above two areas (hereinafter referred to as the area difference) while changing the parameter L within the range of variation. FIG. 8 is a table showing the above calculation results. FIG. 9 is a graph showing the relationship between the parameter L (horizontal axis) and the area difference (vertical axis) shown in FIG. 8 .

[0072] The third threshold value Th3 can be said to be a threshold value for separating the bad mark BM from the background area. Here, referring to FIG. 4, in the example described above, there are mainly bright parts representing the bad mark BM (hereinafter referred to as the bad mark part), dark parts representing the substrate (hereinafter referred to as the substrate part), and parts representing balls brighter than the bad mark BM (hereinafter referred to as the ball part) in the bad mark area ROI. In the image IM1 of FIG. 4, the ball part corresponds to the white dot pattern. Next, referring to FIG. 9, during the period when the parameter L changes from 0 to a certain value V1 (V1>0), the brightness of all pixels usually converges within the range of L~255, so the area difference value is substantially 0. However, if the parameter L exceeds the value V1, the brightness of the pixels in the substrate part begins to deviate from the range of L~255, so the curve of the area difference value becomes steeper. The reason is that due to the presence of the bad mark BM, the substrate part is hidden in the defective area (that is, the previous substrate part is invisible due to the bad mark BM), but the substrate part is not hidden in the good area (that is, the previous substrate part can be seen). That is, when the parameter L exceeds the value V1, the presence of the defective mark BM begins to appear in the form of an area difference.

[0073] If the parameter L exceeds the peak value V2 (V2>V1), the curve of the area difference value drops sharply, and if the parameter L exceeds the value V3 (V3>V2), the curve of the area difference value is gentle or roughly constant. The reason is that the number of pixels in the substrate portion covered by the bad mark BM reaches the peak. Furthermore, the peak value V2 can be said to be a value close to the average brightness of the substrate portion. Thereafter, if the parameter L exceeds the value V4 (V4>V3), the curve of the area difference value begins to drop further. V4 can be said to be a value close to the average brightness of the bad mark BM. Therefore, if the parameter L exceeds the value V4, the brightness of the pixels in the bad mark portion begins to deviate from the range of L~255, so the curve of the area difference value gradually approaches 0. The reason is that the area of the ball portion in the bad mark region ROI in the good product area is the same as that in the bad mark region ROI in the defective product area.

[0074] In the next step S111, the control unit 70 sets the third threshold value Th3 based on the area difference value calculated in step S110. Specifically, the control unit 70 determines the parameter L (hereinafter referred to as the reference parameter Ls) that maximizes the area difference value within the variation range (0-255) of the parameter L. Then, the brightness that differs from the reference parameter Ls by a predetermined offset value F1 is set as the third threshold value Th3 and stored in the storage unit 80. At this time, the value of the reference parameter Ls can also be stored in the storage unit 80 as a reference value. The offset value F1 can be a fixed value or can be automatically set. In the case of automatic setting, for example, the offset value F1 can also be calculated based on the brightness of the pixels in the defective mark area ROI included in the two good product areas and the defective product area that are the object.

[0075] In the example, the third threshold value Th3 is set to a value larger than the offset value F1 with respect to the reference parameter Ls. The reason for this is that the third threshold value Th3 is set within the range of V3 to V4. The reason for this is that the range of V3 to V4 is the part where the number of pixels of the substrate portion covered by the defective mark BM reaches the top, and is the position where the existence of the defective mark BM is stably revealed.

[0076] In the above manner, the defective mark region ROI and the third threshold value Th3 are set as two parameters for defective mark inspection.

[0077] <3-2. Process of defective mark inspection> Next, the process of defective mark inspection is described. During the formal manufacturing of semiconductor component S1, defective mark inspection is performed on the inspection image captured by the second optical inspection camera 13. The inspection image is the same image as the image in the example of FIG. 4. Therefore, the inspection image is also labeled with the symbol IM1 below.

[0078] The control unit 70 continuously acquires the inspection image IM1 and performs the following processing on each inspection image IM1. First, similarly to step S102, the control unit 70 extracts a plurality of component regions T1 to T4 from the inspection image IM1. Then, referring to the storage unit 80, the bad mark region ROI is set in each component region based on the information for determining the position of the bad mark region ROI stored in step S108. Then, referring to the storage unit 80, the third threshold value Th3 is read out. Then, for each bad mark region ROI, it is determined whether the number of pixels contained in the bad mark region ROI that exceeds the third threshold value Th3 is greater than the fourth threshold value Th4. In the case where it is determined that such a number of pixels is greater than the fourth threshold value Th4, it is determined that the bad mark region ROI contains a bad mark BM, and the semiconductor component S1 corresponding to the component region containing the bad mark region ROI is determined to be a defective product. If this is not the case, it is determined that the defective mark region ROI does not include the defective mark BM, and the semiconductor component S1 corresponding to the component region including the defective mark region ROI is determined to be a good product.

[0079] In the above manner, the bad mark inspection is completed. Furthermore, the fourth threshold value Th4 can be a fixed value, or can be set manually or automatically before the bad mark inspection. In the case of automatic setting before the bad mark inspection, for example, the fourth threshold value Th4 can also be calculated based on the brightness of the pixels in the bad mark area ROI included in the two good product areas and the bad product area that are the objects.

[0080] [4. Features] As described above, in the cutting device 1 according to this embodiment, the bad mark region ROI is automatically set in advance. Therefore, the deviation of the setting performed by the operator can be eliminated, and the accuracy of the bad mark inspection is improved. In addition, the setting time required for preparing the bad mark inspection can be shortened.

[0081] In addition, the third threshold value Th3 is also automatically set in advance. Therefore, for the third threshold value Th3, the deviation of the setting performed by the operator can also be eliminated, and the accuracy of the defective mark inspection is further improved. In addition, the setting time required for preparing the defective mark inspection can be further shortened.

[0082] [5. Modifications] The above embodiments are merely examples of the present invention in all aspects. The above embodiments can be variously improved or modified within the scope of the present invention. For example, the following modifications can be made. In the implementation of the present invention, a specific structure can be appropriately adopted according to the embodiments.

[0083] <5-1> In the embodiment described, step S107 may also be omitted.

[0084] <5-2> In the above embodiment, an example is shown in which the bad mark area ROI mainly includes the bad mark part, the substrate part, and the ball part. In addition, in order of brightness, it is set to be the ball part, the bad mark part, and the substrate part. However, even if the bad mark part is the darkest area, the bad mark part and the substrate part are separated near the peak value V2, so the third threshold value Th3 can be calculated by the same algorithm. In this way, the algorithm for calculating the third threshold value Th3 can be applied to various types of semiconductor components S1.

[0085] <5-3> In the above embodiment, the bad mark region ROI and the third threshold value Th3 are automatically set in advance. However, only the bad mark region ROI or only the third threshold value Th3 may be automatically set.

[0086] <5-4> In the above-described embodiment, in step S100 of the preparation step, the same package substrate P1 as that used in the actual manufacturing of the semiconductor component S1 is used, but a dummy substrate of the package substrate P1 may also be used.

[0087] The above is an illustrative description of the embodiment of the present invention. That is, the detailed description and the accompanying drawings are disclosed for the purpose of illustrative description. Therefore, the components described in the detailed description and the accompanying drawings sometimes include components that are not necessary for solving the problem. Therefore, it should not be immediately determined that these non-essential components are necessary just because they are described in the detailed description and the accompanying drawings.

[0088] 1: Cutting device 3: Substrate supply unit 4: Positioning unit 4a: Track Department 5: Cut off the workbench 5a: Retaining member 5b: Rotating mechanism 5c: Mobile mechanism 5d: First position confirmation camera 5e: First Cleaner 6: Spindle 6a:Leaf 6b: Second location confirmation camera 6c: Rotation axis 7:Transportation Department 7a: Second Cleaner 11: Check the workbench 12: First optical inspection camera 13: Second optical inspection camera 14: Configuration 15: Extraction 15a: Good product tray 15b: Pallet for defective products 16, 17: Lighting Department 17a: Dome 17b:LED 20: Monitor 50: Computer 70: Control Department 72:CPU 74: RAM 76:ROM 80: Storage 81: Control Program 90: Input and output I / F 95: Reception Department A1: Cutting module (cutting mechanism) B1: Inspection / storage module (inspection device) BL:Borderline BM: Bad Mark C: Binding area (area) Cmax: Maximum binding area D1~Dn: Division area F1: Offset value IM1: Image (setting image, inspection image) L: Parameters Ls: Benchmark parameters M1: Material box P1: Package substrate ROI: Badly marked area S1: Semiconductor parts S100, S101, S102, S103, S104, S105, S106, S107, S108, S109, S110, S111: Steps T1, T2, T4: Good product area (parts area) T3: Defective product area (parts area) Th1: first threshold Th2: Second threshold Th3: third threshold Th4: fourth threshold V1, V3, V4: value V2: Peak value X, Y, Z: axis θ1: direction

Claims

1. An inspection device, comprising: The control unit, within a preset image containing multiple inspection objects, pre-sets the position of the area (i.e., the defective mark area) where the multiple inspection objects contain defective marks, and inspects the defective mark area within the inspection image to determine whether the multiple inspection objects captured in the inspection image contain the defective mark. The preset image includes multiple part areas corresponding to each of the multiple inspection objects. Pre-setting the position of the defective mark area within the preset image refers to pre-setting the position of the defective mark area within the multiple part areas. The control unit pre-setting the position of the defective mark area within the multiple part areas includes: acquiring the preset image containing the multiple inspection objects, including inspection objects that do not contain the defective mark (i.e., good products) and inspection objects that contain the defective mark (i.e., defective products). The brightness of at least one of the plurality of part areas (i.e., good product areas) containing the good product in the set image is compared with the brightness of at least one of the plurality of part areas (i.e., defective product areas) containing the defective product in the set image, thereby determining the defective marking area within the plurality of part areas.

2. The inspection apparatus as claimed in claim 1, wherein, The control unit divides each of the good product area and the defective product area into multiple segmented areas. For each segmented area, the brightness of the good product area is compared with the brightness of the defective product area to determine the defective marking area.

3. The inspection apparatus as described in claim 2, wherein, The control unit selects the segmented region where the brightness difference between the good product region and the defective product region is greater than or equal to a first threshold value, and determines the defective marking region in a manner that includes the selected segmented region.

4. The inspection apparatus as described in claim 3, wherein, In addition to selecting the segmented regions where the brightness difference is above the first threshold, the control unit further selects the segmented regions adjacent to the segmented regions where the brightness difference is above the first threshold and the segmented regions where the brightness difference is above the second threshold, and determines the defective marking region in a manner that includes the selected segmented regions.

5. The inspection apparatus as described in claim 3 or 4, wherein, The control unit determines the defective marking region in a manner that includes the region with the largest area among one or more regions formed by connecting the selected segmented regions.

6. The inspection device as claimed in any one of claims 1 to 4, wherein, The control unit presets a third threshold value for inspection of the defective marking area in the inspection image based on the brightness of the defective marking area in the setting image.

7. The inspection apparatus as claimed in claim 6, wherein, The third threshold value can be preset in the following ways: while changing the third threshold value, the area of ​​the portion of the defective marking area contained in the good product area whose brightness exceeds the third threshold value is calculated and the area of ​​the portion of the defective marking area contained in the defective product area whose brightness exceeds the third threshold value is calculated, and the third threshold value is preset based on the difference.

8. The inspection apparatus as claimed in claim 7, wherein, The control unit presets the brightness of the third threshold value as the value offset from the third threshold value that maximizes the difference within the variation range of the third threshold value.

9. A cutting device, comprising: The cutting mechanism cuts a substrate containing multiple semiconductor components and monolithically integrates the semiconductor components. And an inspection apparatus as described in any one of claims 1 to 8, which inspects the monolithized semiconductor component as one of the plurality of inspection objects.

10. A method for manufacturing a semiconductor component, comprising the steps of: cutting a substrate comprising a plurality of semiconductor components by means of a cutting mechanism and monolithizing the semiconductor components; and inspecting the monolithized semiconductor component as one of the plurality of inspection objects using an inspection apparatus as described in any one of claims 1 to 8.

Images