Chip opening method and opening device based on machine vision

Abstract

The application relates to the chip opening technology field, in particular to a chip opening method based on machine vision, which comprises the following steps: emitting an opening laser beam to the surface of a chip to open the chip layer by layer. Before and after the chip opening, an indicating laser line is projected, a structured light image of the indicating laser line is collected, the vertical distance between the indicating laser light source and the indicating laser line is calculated, and the opening depth is calculated by using the vertical distances before and after. Whether to continue opening is confirmed according to the opening depth. The application also relates to an opening device based on machine vision, which comprises an opening mechanism, an indicating laser light source, an image collection device, a visual imaging mechanism and a processor. The processor calculates the opening depth according to the collected image, and sends an opening start signal or an opening stop signal to the opening mechanism in combination with the image data of the visual imaging mechanism. Under the premise of layer-by-layer opening of the chip, the opening process is closed-loop regulated.

Description

Technical Field

[0001] This application relates to the field of chip unpacking technology, and more specifically, to a chip unpacking method and unpacking device based on machine vision. Background Technology

[0002] Laser decapsulation machines use high-power, high-energy laser beams to sweep and remove the surface material of pre-packaged integrated circuit chips. Following a preset trajectory, the surface material is physically etched layer by layer until the internal metal contact structure of the chip is exposed. This prepares the material for subsequent chip failure analysis, allowing testing personnel to easily test the chip's reliability and performance. The decapsulation process must not damage the chip's functionality and must keep the chip, bonding pads, bonding wires, and lead frames in an undamaged state.

[0003] Existing chip unpacking technologies have the following problems: Chip unpacking methods use laser emission to unpack the chip along a preset trajectory. During the unpacking process, the chip unpacking machine can only unpack according to a given value and cannot detect and provide feedback on the unpacking depth, which can lead to under-unpacking or over-unpacking, damaging the chip.

[0004] Therefore, there is a need to provide an improved technical solution that addresses the shortcomings of the existing technology. Summary of the Invention

[0005] The purpose of this application is to provide a chip unpacking method and unpacking device based on machine vision, which enables closed-loop regulation of the unpacking process while unpacking the chip layer by layer.

[0006] Firstly, a machine vision-based chip unpacking method is provided, comprising the following steps:

[0007] S1, project a first indicator laser line onto the surface of the micro electronic device, acquire the structured light image of the first indicator laser line, and calculate the vertical distance L1 between the indicator laser source and the first indicator laser line.

[0008] S2 uses an opening laser beam to open the surface of the microelectronic device.

[0009] S3, project a second indicator laser line onto the surface of the opened micro electronic device, acquire a structured light image of the second indicator laser line, and calculate the vertical distance L2 between the indicator laser source and the second indicator laser line.

[0010] S4, use L1 and L2 to calculate the opening depth z.

[0011] S5, determine whether the opening depth z is less than the set depth h; if the opening depth z is less than the set depth h, repeat steps S3-S4; stop opening when the opening depth z is equal to the set depth h.

[0012] In one embodiment, a microelectronic device is placed on a stage; an image acquisition device is used to acquire a structured light image of an indicator laser line; wherein...

[0013] The structured light image of the indicator laser line is a tomographic image formed between the indicator laser line on the upper surface of the microelectronic device and the upper surface of the stage.

[0014] In one embodiment, the calculation method of L1 includes: obtaining the baseline distance AB from the indicator laser source A to the center point B of the lens of the image acquisition device, obtaining the limit displacement distance x of the indicator laser source in the image coordinate system, and obtaining the camera focal length f of the image acquisition device.

[0015] Formula 1 is derived from the similarity theorem of triangles: The vertical distance m1 from the fracture point O to AB of the first indicator laser line is calculated according to Formula 1.

[0016] Obtain the angle α between L1 and AB using the formula. L1 is calculated.

[0017] The calculation method for L2 includes: obtaining the distance AB from the indicator laser source A to the center point B of the lens of the image acquisition device, obtaining the limit displacement distance y of the indicator laser source in the image coordinate system, and obtaining the camera focal length f of the image acquisition device.

[0018] Formula 2 is derived from the similarity theorem of triangles: Calculate the distance L2 from the indicator laser source A to the fracture point O' of the second indicator laser line according to Formula 2.

[0019] To obtain the angle α between L2 and AB, use the formula... L2 is calculated.

[0020] In one implementation, in S4, calculating the opening depth z using L1 and L2 includes: calculating the opening depth z using the formula z = L2 - L1.

[0021] In one embodiment, in S1 and S3, acquiring structured light images of the indicator laser line includes: acquiring multiple structured light images of the indicator laser line along the direction perpendicular to the indicator laser line; in S4, calculating the opening depth z includes: calculating multiple opening depths z based on the multiple structured light images of the indicator laser line, and using the average value of the multiple opening depths z as the opening depth value.

[0022] In one implementation, the image acquisition speed of the image acquisition device is 20-40 frames per second.

[0023] In one implementation, in S2, the feed amount for opening the microelectronic device in one operation is γ. In S5, when determining whether the opening depth z is less than the set depth h, the method further includes: determining whether the difference between the opening depth z and the set depth h is less than the feed amount γ; if the difference between the opening depth z and the set depth h is less than the feed amount γ, the feed amount γ for opening is adjusted.

[0024] In one embodiment, a feature identifier is provided on the surface of the microelectronic device, and different opening procedures are matched for different feature identifiers. Before S1, the method further includes: collecting the feature identifier of the microelectronic device and selecting the opening procedure corresponding to the microelectronic device according to the feature identifier.

[0025] In one implementation, the method further includes: acquiring images of the microelectronic device in real time, extracting the edge grayscale values ​​of the microelectronic device images and comparing them with preset values, and adjusting the output power value of the Kaifeng laser based on the comparison results.

[0026] According to a second aspect of this application, a machine vision-based chip unpacking device is also provided, comprising:

[0027] The opening mechanism includes a fiber laser for projecting the opening laser.

[0028] Indicator laser source, used to project an indicator laser line;

[0029] An image acquisition device for acquiring structured light images of the indicator laser line emitted by an indicator laser source;

[0030] Visual imaging mechanisms extract image data through structured light imaging;

[0031] The processor is connected in communication with the image acquisition device, the vision imaging mechanism, and the opening mechanism. It is used to calculate the opening depth z based on the image acquired by the image acquisition device, and send an opening start signal or an opening stop signal to the opening mechanism in combination with the image data from the vision imaging mechanism.

[0032] The beneficial effects of the machine vision-based chip unpacking method and device described in this application are as follows:

[0033] By unpacking the chip layer by layer, the unpacking depth is calculated after each layer is unpacked, and the unpacking depth can be obtained in real time. The unpacking depth is compared with the set depth to determine whether to continue unpacking.

[0034] Before opening the package, the corresponding opening procedure is extracted based on the feature identifiers of the machine vision acquisition chip, thus avoiding operational errors caused by human selection of the procedure.

[0035] The difference between the opening depth and the set depth is compared with the opening feed rate. Based on the comparison, the opening feed rate γ is adjusted in real time. This allows for real-time detection and feedback of the opened depth, adjusting the opening feed rate to avoid under- or over-opening and damaging the chip. This closed-loop control improves opening accuracy and prevents chip damage during the opening process.

[0036] By acquiring real-time images of the chip after each layer is opened, extracting the edge grayscale values ​​of the images, and comparing them with preset values, the output power of the opening laser is adjusted based on the comparison results. This allows for ablation detection based on machine vision and closed-loop control of the opening laser's output power. Attached Figure Description

[0037] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart illustrating a machine vision-based chip unpacking method according to an embodiment of this application;

[0039] Figure 2 A schematic diagram indicating a laser line scanning chip;

[0040] Figure 3 A schematic diagram for calculating L1 and L2;

[0041] Figure 4 This is a schematic diagram of a machine vision-based chip unpacking device according to an embodiment of this application.

[0042] Figure label:

[0043] 1-Fiber laser; 2-Image acquisition device; 3-Processor; 4-Chip. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of the embodiments of this application clearer, the technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of this application, and not all embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations.

[0045] Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely to illustrate selected embodiments of the application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0046] Figure 1 This is a flowchart illustrating a machine vision-based chip unpacking method according to an embodiment of this application. See also... Figure 1 This invention provides a chip unpacking method based on machine vision, comprising the following steps:

[0047] S1, a first indicator laser line is projected onto the chip surface, and an image acquisition device acquires a structured light image of the first indicator laser line and calculates the vertical distance L1 from the indicator laser source A to the first indicator laser line.

[0048] S2 uses an opening laser beam to be emitted onto the surface of the chip to open the chip once.

[0049] S3. After one unpacking is completed, a second indicator laser line is projected onto the surface of the unpacked chip. The image acquisition device acquires the structured light image of the second indicator laser line and calculates the vertical distance L2 from the indicator laser source A to the second indicator laser line.

[0050] S4, use L1 and L2 to calculate the opening depth z.

[0051] S5, determine whether the opening depth z is less than the set depth h; if the current opening depth z is less than the set depth h, repeat steps S3-S4; stop opening when the current opening depth z is equal to the set depth h.

[0052] In one embodiment, in step S1, the chip is placed on a stage, and an image acquisition device is used to acquire a structured light image of the indicator laser line. The structured light image of the indicator laser line is a tomographic image formed by the indicator laser line between the upper surface of the chip and the upper surface of the stage. Figure 2 As shown, due to the height difference between the upper surface of the chip and the upper surface of the stage, the tomography occurs at the edge of the chip. This results in extremely high recognition rate when acquiring structured light images, improving the visualization effect, and also provides a significant color contrast between the chip and the stage. Therefore, tomographic images are used to calculate L1 and L2.

[0053] In one implementation scheme, such as Figure 3 As shown, the calculation method for L1 includes:

[0054] According to the similarity theorem of triangles:

[0055]

[0056] In the formula:

[0057] AB represents the baseline distance from laser source A to the center point B of the lens of the image acquisition device, which is obtained through calibration.

[0058] DE is the limit displacement distance of the laser source in the image coordinate system. It is a known parameter and is denoted as x.

[0059] BF is the camera focal length of the image acquisition device, a known parameter, denoted as f;

[0060] m1 is the vertical distance from the fracture point O to AB of the first indicator laser line.

[0061] From equation (1), we can derive:

[0062]

[0063] L1 is the distance from the indicator laser source A to the fracture point O of the first indicator laser line:

[0064]

[0065] In the formula:

[0066] α is the angle between AO and AB, which is determined by the image acquisition device.

[0067] From equation (2) and equation (3), we can derive:

[0068]

[0069] In one implementation scheme, such as Figure 3 As shown, the calculation method for L2 includes:

[0070] According to the similarity theorem of triangles:

[0071]

[0072] In the formula:

[0073] AB represents the distance from laser source A to the center point B of the lens of the image acquisition device, which is obtained through calibration.

[0074] DG is the limit displacement distance of the laser source in the image coordinate system. The parameter is known and denoted as y.

[0075] BF is the camera focal length of the image acquisition device, a known parameter, denoted as f;

[0076] m2 is the vertical distance from the fracture point O' of the second indicator laser line to AB.

[0077] From equation (5), we can derive:

[0078]

[0079] L2 is the distance from the indicator laser source A to the fracture point O' of the second indicator laser line:

[0080]

[0081] In the formula:

[0082] α is the angle between AO and AB, which is determined by the image acquisition device.

[0083] From equation (6) and equation (7), we can derive:

[0084]

[0085] In one implementation scheme, such as Figure 2 As shown, the image acquisition device calculates the parallax relative pixel distance based on the indicator laser line of the staggered fault and the focal plane characteristics of virtual confocal imaging. This distance can be reflected by transforming the spatial coordinates of each point in the image to reflect the spatial opening depth z. In S4, the calculation of the opening depth z using L1 and L2 includes: calculating the opening depth z using the formula z = L2 - L1.

[0086] In one implementation scheme, such as Figure 2 As shown, in S1 and S3, the direction of the indicator laser line is the X direction, and the direction perpendicular to the indicator laser line is the Y direction. Acquiring the structured light image of the indicator laser line includes: moving the chip or the indicator laser line along the Y direction, continuously acquiring multiple indicator laser images using an image acquisition device, generating a structured light image of the indicator laser line of the fault based on the laser line scanning imaging algorithm, and combining the structured light image and the machine vision algorithm to obtain a three-dimensional point cloud map, so as to more accurately confirm the location of the fault.

[0087] In S4, in order to reduce measurement numerical errors, the calculation of the opening depth z includes: calculating multiple opening depths z based on multiple structured light images of the indicator laser lines, and using the average value of the multiple opening depths z as the opening depth value.

[0088] In one implementation, the limit displacement distance DE or DG of the laser spot in the image coordinate system is calculated from the pixels of the image. Given that the pixel size is 10μm, the number of pixels covered by the straight line of DE or DG is calculated, and the two are multiplied to obtain the physical value x of DE and the physical value y of DG.

[0089] In one embodiment, the image acquisition device is a CCD camera, and the image acquisition speed of the CCD camera is 30 frames per second.

[0090] In one implementation, in step S2, the feed amount γ for opening the chip is 0.1 mm. In step S5, after determining whether the opening depth z is less than the set depth h, the method further includes: determining whether the difference between the opening depth z and the set depth h is less than the feed amount γ; if the difference between the opening depth z and the set depth h is less than the feed amount γ, the opening feed amount γ is adjusted to 0.01 mm. Real-time adjustment of the feed amount γ achieves closed-loop control, improves the accuracy of opening, and avoids damage to the chip during opening.

[0091] In one implementation, a feature identifier is provided on the chip surface. The feature identifier includes contour features and OCR features. Different feature identifiers are matched with different unpacking programs. Before S1, the method further includes: acquiring the feature identifiers on the chip surface and selecting the unpacking program corresponding to the chip based on the feature identifiers. The setting parameters in the unpacking program include the output power value of the unpacking laser, the set unpacking depth h of the chip, and the feed amount γ for each unpacking layer. This method can reverse-match the unpacking program based on machine vision. Before unpacking, the feature identifiers of the chip are acquired based on machine vision algorithms, and the corresponding unpacking program is extracted, avoiding operational errors caused by manual program selection.

[0092] In one implementation, to prevent excessive burning, the method further includes: real-time acquisition of chip images, extraction of edge grayscale values ​​from the chip images and comparison with preset values, and adjustment of the output power of the opening laser based on the comparison results. The chip images are acquired using an image acquisition device, binarized, and edge grayscale values ​​(RGB = 0 for white and RGB = 255 for black) are extracted. When the edge grayscale value (RGB) is greater than or equal to 75, the burning mechanism is triggered, and a hexadecimal character is sent to the host computer via the RS485 protocol to reduce the output power of the opening laser. This method enables closed-loop control of the opening laser's output power based on machine vision-based burning detection.

[0093] This invention also provides a chip unpacking device based on machine vision, such as... Figure 4 As shown, it includes:

[0094] The opening mechanism includes a fiber laser 1 for projecting an opening laser.

[0095] Indicator laser light source, used to project an indicator laser line.

[0096] Image acquisition device 2 is used to acquire structured light images of the indicator laser line emitted by the indicator laser source and images of the acquisition chip 4.

[0097] The visual imaging mechanism extracts image data from the image on chip 4.

[0098] The processor 3 is communicatively connected to the image acquisition device, the visual imaging mechanism, and the opening mechanism. It is used to calculate the opening depth z based on the image acquired by the image acquisition device, and send an opening start signal or an opening stop signal to the opening mechanism in combination with the image data of the visual imaging mechanism.

[0099] This invention provides a chip unpacking method and device based on machine vision. By unpacking the chip layer by layer, the unpacking depth is calculated after each layer is unpacked, so as to obtain the real-time unpacking depth. The unpacking depth is compared with the set depth to determine whether to continue unpacking.

[0100] Before opening the package, this invention extracts the corresponding opening procedure based on the feature identifier of the machine vision acquisition chip, thus avoiding operational errors caused by human selection of the procedure.

[0101] The difference between the opening depth and the set depth is compared with the opening feed rate. Based on the comparison, the opening feed rate γ is adjusted in real time. This allows for real-time detection and feedback of the opened depth, adjusting the opening feed rate to avoid under- or over-opening and damaging the chip. This closed-loop control improves opening accuracy and prevents chip damage during the opening process.

[0102] This invention acquires images of the chip after each layer is opened in real time, extracts the edge grayscale values ​​of the images and compares them with preset values, and adjusts the output power of the opening laser based on the comparison results. It can perform ablation detection based on machine vision and implement closed-loop control of the opening laser's output power.

[0103] The above description is merely a preferred embodiment of this application and is not intended to limit this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the protection scope of this application.

Claims

1. A chip unpacking method based on machine vision, characterized in that, Includes the following steps: S1, Project a first indicator laser line onto the surface of the micro electronic device, acquire a structured light image of the first indicator laser line, and calculate the vertical distance L1 between the indicator laser source and the first indicator laser line; S2, using an opening laser beam emitted onto the surface of the microelectronic device to open the microelectronic device once; the feed amount for opening the microelectronic device once is γ; S3, project a second indicator laser line onto the surface of the opened micro electronic device, acquire a structured light image of the second indicator laser line, and calculate the vertical distance L2 between the indicator laser source and the second indicator laser line; In S1 and S3, the microelectronic device is placed on a stage and a structured light image is acquired using an image acquisition device; the acquired structured light image is a tomographic image formed between the upper surface of the microelectronic device and the upper surface of the stage by the first or second indicator laser line. S4, calculate the opening depth z using L1 and L2; S5, determine whether the opening depth z is less than the set depth h; if the opening depth z is less than the set depth h, repeat steps S3-S4; stop opening when the opening depth z equals the set depth h; When determining whether the opening depth z is less than the set depth h, the method further includes: determining whether the difference between the opening depth z and the set depth h is less than the feed amount γ; if the difference between the opening depth z and the set depth h is less than the feed amount γ, the opening feed amount γ is adjusted. After each layer is opened, images of the microelectronic devices are acquired in real time, the edge grayscale values ​​of the microelectronic device images are extracted and compared with preset values, and the output power value of the opening laser is adjusted according to the comparison results.

2. The chip unpacking method based on machine vision according to claim 1, characterized in that, The calculation method for L1 includes: Obtain the baseline distance AB from the indicator laser source A to the center point B of the lens of the image acquisition device; Obtain the limit displacement distance x of the indicator laser source in the image coordinate system; Obtain the camera focal length f of the image acquisition device; Formula 1 is derived from the similarity theorem of triangles: The vertical distance m1 from the fracture point O to AB of the first indicator laser line is calculated according to Formula 1. Find the angle α between L1 and AB; According to the formula The L1 is calculated; The calculation method for L2 includes: Obtain the baseline distance AB from the indicator laser source A to the center point B of the lens of the image acquisition device; Obtain the limit displacement distance y of the indicator laser source in the image coordinate system; Obtain the camera focal length f of the image acquisition device; Formula 2 is derived from the similarity theorem of triangles: Calculate the distance L2 from the indicator laser source A to the fracture point O' of the second indicator laser line according to Formula 2; Find the angle α between L2 and AB; According to the formula The L2 is calculated.

3. The chip unpacking method based on machine vision according to claim 1, characterized in that, In S4, calculating the opening depth z using L1 and L2 includes: using the formula The opening depth z is calculated.

4. The chip unpacking method based on machine vision according to claim 1, characterized in that, In S1 and S3, acquiring structured light images of the indicator laser line includes: acquiring multiple structured light images of the indicator laser line along a direction perpendicular to the indicator laser line; In S4, calculating the opening depth z includes: calculating multiple opening depths z based on multiple structured light images of the indicator laser lines, and using the average of the multiple opening depths z as the opening depth value.

5. The chip unpacking method based on machine vision according to claim 1, characterized in that, The image acquisition speed of the image acquisition device is 20-40 images per second.

6. The chip unpacking method based on machine vision according to claim 1, characterized in that, The surface of the micro electronic device is provided with a feature identifier, and different opening procedures are matched for different feature identifiers. Before S1, the method further includes: collecting the feature identifier of the micro electronic device, and selecting the opening procedure corresponding to the micro electronic device according to the feature identifier.

7. A chip unpacking apparatus employing the chip unpacking method as described in any one of claims 1-6, characterized in that, include: The opening mechanism includes a fiber laser for projecting the opening laser. The microelectronic device is then opened layer by layer according to the opening feed rate γ; Indicator laser source, used to project an indicator laser line; Image acquisition device, used to acquire structured light images of the indicator laser line emitted by the indicator laser source and images of the microelectronic devices after each layer is opened; A visual imaging mechanism extracts image data from the structured light image and the microelectronic device image; The processor is communicatively connected to the image acquisition device, the visual imaging mechanism, and the opening mechanism. It is used to calculate the vertical distance L1 between the indicator laser source and the first indicator laser line based on the structured light image of the first indicator laser line acquired before opening, calculate the vertical distance L2 between the indicator laser source and the second indicator laser line based on the structured light image of the second indicator laser line acquired after one opening, and calculate the opening depth z using L1 and L2. The processor is further configured to compare the opening depth z with the set depth h, and when the opening depth z is less than the set depth h, determine whether the difference between the set depth h and the opening depth z is less than the opening feed amount γ, and when the difference is less than the opening feed amount γ, control the opening mechanism to adjust the opening feed amount γ. The processor is also used to control the unpacking mechanism to adjust the output power value of the fiber laser based on the comparison result between the edge gray value of the micro electronic device image after each layer of unpacking extracted by the visual imaging mechanism and the preset value. When the opening depth z reaches the set depth h, the processor sends an opening stop signal to the opening mechanism.

Images