# US 2021/0242059 A1 · System and Method for Semiconductor Device Print Check Alignment > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 2021/0242059 A1 - Document ID: US20210242059A1 - Status: Published application - Kind: A1 - Application number: 16/778,607 - Filed: 2020-01-31 - Published: 2021-08-05 - Earliest claimed: 2020-01-31 - Assignees: KLA Corporation - Applicants: KLA Corporation - Inventors: Hong Chen - Web page: https://patentgrep.com/patent/US20210242059A1 ## CPC - H10P 72/53: {using optical controlling means} - G06T 7/001: {using an image reference approach} - G06V 10/751: {Comparing pixel values or logical combinations thereof, or feature values having positional relevance, e.g. template matching} - G06V 10/7515: {Shifting the patterns to accommodate for positional errors} - G06V 10/774: Generating sets of training patterns; Bootstrap methods, e.g. bagging or boosting - G06V 20/52: Surveillance or monitoring of activities, e.g. for recognising suspicious objects (recognising microscopic objects G06V20/69) - H10P 72/06: {Apparatus for monitoring, sorting, marking, testing or measuring} - H10P 72/0606: {Position monitoring, e.g. misposition detection or presence detection} - G06T 2207/20021: Dividing image into blocks, subimages or windows - G06T 2207/30148: Semiconductor; IC; Wafer - G06V 2201/06: Recognition of objects for industrial automation - H10P: GENERIC PROCESSES OR APPARATUS FOR THE MANUFACTURE OR TREATMENT OF DEVICES COVERED BY CLASS H10 - G06T: IMAGE DATA PROCESSING OR GENERATION, IN GENERAL - G06V: IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING ## Abstract A system includes a controller communicatively coupled to an optical inspection sub-system, the controller configured to: receive training images of a sample; identify alignment targets in the master die; receive a first set of reference images of a first die row of a sample, the first die row including a master die and a first set of reference die; align the first set of reference die to the master die via fine alignment processes to generate a first set of aligned reference images of the first row; receive a second set of reference images of a second die row of the sample; align the second set of reference die to the master die based on the alignment targets and the training images using a course alignment offset value; and align the second set of reference die to the master die via a fine alignment process. ## Claims ### Claim 1 A system, comprising: a controller communicatively coupled to an optical inspection sub-system, the controller including one or more processors configured to execute a set of program instructions stored in memory, the set of program instructions configured to cause the one or more processors to: receive one or more training images of a sample, the sample including a master die; identify one or more alignment targets in the master die and store the one or more alignment targets in memory; receive a first set of one or more reference images of a first die row of a sample, the first die row including a master die and a first set of one or more reference die; align the first set of one or more reference die to the master die via one or more fine alignment processes to generate a first set of one or more aligned reference images of the first row; receive a second set of one or more reference images of a second die row of the sample, the second die row including a second set of one or more reference die; align the second set of one or more reference die to the master die based on the alignment targets and the one or more training images using a course alignment offset value; and align the second set of one or more reference die to the master die via a fine alignment process to generate a second set of one or more aligned reference images of the second die row. ### Claim 2 (depends on 1) The system in claim 1, determine one or more characteristics of the sample based on one or more differences between the one or more training images and at least one of the first set of one or more aligned reference images or the second set of one or more aligned reference images. ### Claim 3 (depends on 2) The system in claim 2, wherein the one or more characteristics of the sample comprise one or more defects of the sample. ### Claim 4 (depends on 1) The system in claim 1, wherein the controller is configured to determine the coarse alignment offset value by: determining an offset value between the second set of one or more reference images and the one or more training images for each of the one or more alignment targets; calculating a normalized cross correlation (NCC) value for each of the one or more alignment targets; generating a sorted list of the one or more alignment targets according to the associated NCC values from largest to smallest; identifying a cluster of alignment targets in the sorted list which exhibit offset values within a selected range of one another; and determining the coarse alignment offset value based on the offset values of the identified cluster of alignment targets. ### Claim 5 (depends on 4) The system in claim 4, wherein determining the coarse alignment offset value based on the offset values of the identified cluster of alignment targets comprises: determining at least one of a median or a mean of the offset values of the identified cluster of alignment targets. ### Claim 6 (depends on 4) The system in claim 4, wherein identifying a cluster of alignment targets in the sorted list which exhibit offset values within a selected range of one another comprises: identifying a two or more clusters of alignment targets in the sorted list which exhibit offset values within a selected range of one another; and identifying the cluster of alignment targets with the largest average NCC value. ### Claim 7 (depends on 1) The system in claim 1, wherein the controller is configured to determine the coarse alignment offset value by: determining an offset value between the second set of one or more reference images and the one or more training images for each of the one or more alignment targets; calculating a sum squared difference (SSD) value for each of the one or more alignment targets; generating a sorted list of the one or more alignment targets according to the associated SSD values from smallest to largest; identifying a cluster of alignment targets in the sorted list which exhibit offset values within a selected range of one another; and determining the coarse alignment offset value based on the offset values of the identified cluster of alignment targets. ### Claim 8 (depends on 1) The system in claim 1, wherein the first set of one or more aligned reference images of the first row are stored in memory. ### Claim 9 (depends on 1) The system in claim 1, wherein the second set of one or more aligned reference images of the second row are stored in memory. ### Claim 10 (depends on 1) The system in claim 1, wherein the controller is further configured to: receive at least one additional set of one or more reference images of at least one additional die row of the sample, the at least one additional die row including at least one additional set of one or more reference die; align the at least one additional set of one or more reference die to the master die based on the alignment targets and the one or more training images using a course alignment offset value; and align the at least one additional set of one or more reference die to the master die via a fine alignment process to generate at least one additional set of one or more aligned reference images of the at least one additional die row. ### Claim 11 (depends on 10) The system in claim 10, wherein the coarse alignment offset value used to align the second die row is different from the coarse alignment offset value used to align the at least one additional die row. ### Claim 12 (depends on 1) The system in claim 1, wherein at least one of the one or more training images, the first set of one or more reference images, or the second set of one or more reference images are received from the optical inspection sub-system. ### Claim 13 (depends on 1) The system in claim 1, wherein the one or more training images are acquired during a scan of a first sample, wherein the first set of one or more reference images are acquired during a first swath scan of a second wafer, and wherein the second set of one or more reference images are acquired during a second swath scan of the second wafer. ### Claim 14 (depends on 1) The system in claim 1, wherein the controller is configured to align the second set of one or more reference die to the master die via a fine alignment process by measuring a plurality of individual offsets of the one or more alignment targets between the second set of one or more reference images and the one or more training images. ### Claim 15 (depends on 14) The system in claim 14, wherein the controller is configured to align the second set of one or more reference die to the master die via a fine alignment process by adjusting the measured plurality of individual offsets to correct distortion. ### Claim 16 A system, comprising: an optical inspection sub-system configured to acquire images of a sample; a controller communicatively coupled to the optical inspection sub-system, the controller including one or more processors configured to execute a set of program instructions stored in memory, the set of program instructions configured to cause the one or more processors to: receive one or more training images of a first sample from the optical inspection sub-system, the first sample including a master die; identify one or more alignment targets in the master die and store the one or more alignment targets in memory; receive a first set of one or more reference images of a first die row of a second sample from the optical inspection sub-system, the first die row including a master die and a first set of one or more reference die; align the first set of one or more reference die to the master die via one or more fine alignment processes to generate a first set of one or more aligned reference images of the first row; receive a second set of one or more reference images of a second die row of the second sample from the optical inspection sub-system, the second die row including a second set of one or more reference die; align the second set of one or more reference die to the master die based on the alignment targets and the one or more training images using a course alignment offset value; and align the second set of one or more reference die to the master die via a fine alignment process to generate a second set of one or more aligned reference images of the second die row. ### Claim 17 (depends on 16) The system in claim 16, determine one or more characteristics of the second sample based on one or more differences between the one or more training images and at least one of the first set of one or more aligned reference images or the second set of one or more aligned reference images. ### Claim 18 (depends on 17) The system in claim 17, wherein the one or more characteristics of the second sample comprise one or more defects of the second sample. ### Claim 19 (depends on 16) The system in claim 16, wherein the controller is configured to determine the coarse alignment offset value by: determining an offset value between the second set of one or more reference images and the one or more training images for each of the one or more alignment targets; calculate a normalized cross correlation (NCC) value for each of the one or more alignment targets; generate a sorted list of the one or more alignment targets according to the associated NCC values from largest to smallest; identify a cluster of alignment targets in the sorted list which exhibit offset values within a selected range of one another; and determine the coarse alignment offset value based on the offset values of the identified cluster of alignment targets. ### Claim 20 (depends on 19) The system in claim 19, wherein determining the coarse alignment offset value based on the offset values of the identified cluster of alignment targets comprises: determining at least one of a median or a mean of the offset values of the identified cluster of alignment targets. ### Claim 21 (depends on 19) The system in claim 19, wherein identifying a cluster of alignment targets in the sorted list which exhibit offset values within a selected range of one another comprises: identifying a two or more clusters of alignment targets in the sorted list which exhibit offset values within a selected range of one another; and identifying the cluster of alignment targets with the largest average NCC value. ### Claim 22 (depends on 16) The system in claim 16, wherein the controller is configured to determine the coarse alignment offset value by: determining an offset value between the second set of one or more reference images and the one or more training images for each of the one or more alignment targets; calculate a sum squared difference (SSD) value for each of the one or more alignment targets; generate a sorted list of the one or more alignment targets according to the associated SSD values from largest to smallest; identify a cluster of alignment targets in the sorted list which exhibit offset values within a selected range of one another; and determine the coarse alignment offset value based on the offset values of the identified cluster of alignment targets. ### Claim 23 (depends on 16) The system in claim 16, wherein the controller is further configured to: receive at least one additional set of one or more reference images of at least one additional die row of the second sample from the optical inspection sub-system, the at least one additional die row including at least one additional set of one or more reference die; align the at least one additional set of one or more reference die to the master die based on the alignment targets and the one or more training images using a course alignment offset value; and align the at least one additional set of one or more reference die to the master die via a fine alignment process to generate at least one additional set of one or more aligned reference images of the at least one additional die row. ### Claim 24 (depends on 23) The system in claim 23, wherein the coarse alignment offset value used to align the second die row is different from the coarse alignment offset value used to align the at least one additional die row. ### Claim 25 (depends on 16) The system in claim 16, wherein at least one of the one or more training images, the first set of one or more reference images, or the second set of one or more reference images are received from the optical inspection sub-system. ### Claim 26 (depends on 16) The system in claim 16, wherein the controller is configured to align the second set of one or more reference die to the master die via a fine alignment process by measuring a plurality of individual offsets of the one or more alignment targets between the second set of one or more reference images and the one or more training images. ### Claim 27 (depends on 26) The system in claim 26, wherein the controller is configured to align the second set of one or more reference die to the master die via a fine alignment process by adjusting the measured plurality of individual offsets to correct distortion. ### Claim 28 A method, comprising: acquiring one or more training images of a sample, the sample including a master die; identifying one or more alignment targets in the master die and storing the one or more alignment targets in memory; acquiring a first set of one or more reference images of a first die row of a sample, the first die row including a master die and a first set of one or more reference die; aligning the first set of one or more reference die to the master die via one or more fine alignment processes to generate a first set of one or more aligned reference images of the first row; acquiring a second set of one or more reference images of a second die row of the sample, the second die row including a second set of one or more reference die; aligning the second set of one or more reference die to the master die based on the alignment targets and the one or more training images using a course alignment offset value; and aligning the second set of one or more reference die to the master die via a fine alignment process to generate a second set of one or more aligned reference images of the second die row.