Patterned wafer microstructure data acquisition system in bright field environment and data set construction method

Abstract

The invention discloses a patterned wafer microstructure data acquisition system in a bright field environment and a data set construction method. The system comprises an adjustable light source assembly, a microscope imaging assembly and a three-axis electric displacement platform, the data set construction comprises the following steps: fixing a patterned wafer sample on a three-axis electric displacement platform, and setting illumination brightness through an adjustable light source assembly to obtain uniform bright field illumination; the three-axis electric displacement platform is controlled to move along a preset path, automatic focusing scanning in the Z-axis direction is executed at each point position, and an optimal focal plane image is generated; the collected images are classified after being cut and preprocessed, and normal samples and abnormal samples are distinguished; for the abnormal sample, generating a defect mask file through interactive labeling; and integrating images and masks to form a structured data set which can be directly used for algorithm research. According to the method, microscopic images with high depth of field and high contrast can be obtained, and a high-quality anomaly detection data set directly used for deep learning model training is constructed.

Description

Technical Field

[0001] This invention relates to the interdisciplinary field of semiconductor inspection, computer vision and industrial optical imaging, and in particular to a data acquisition system and dataset construction method for patterned wafer microstructures under bright field conditions. Background Technology

[0002] As the semiconductor industry moves towards smaller feature sizes, higher integration densities, and more complex pattern structures, the importance of wafer surface defect detection in integrated circuit manufacturing is becoming increasingly prominent.

[0003] In patterned wafer manufacturing, any minute structural anomalies (such as particle contamination, impurity residue, or process defects) can lead to device malfunction, severely impacting product yield and reliability. Therefore, high-precision, high-efficiency wafer defect detection technology has become a critical aspect of manufacturing process control.

[0004] Currently, there are two main types of defect detection technologies for patterned wafers in industry: electron beam inspection and optical inspection. Electron beam inspection has extremely high spatial resolution and can identify nanoscale defects, making it an effective means of achieving ultra-fine feature defect analysis. However, this technology suffers from slow inspection speed and high equipment costs, making it difficult to meet the needs of large-area inspection. In contrast, optical inspection methods, especially bright-field inspection, remain the mainstream inspection solution in the mass production of patterned wafers due to their advantages such as high throughput, non-destructive nature, and lower cost.

[0005] However, traditional bright-field detection systems still have the following problems: (1) Uneven illumination and reflection interference: Under complex pattern structures, specular reflection can easily lead to local overexposure or underexposure, weakening the visibility of defects; (2) Insufficient depth of field in single focal plane imaging: There are small height differences on the wafer surface, and a single focal plane cannot simultaneously ensure global clarity; (3) Low data annotation efficiency and poor consistency: Automatic acquisition and manual annotation processes are separated, and defect identification relies on manual experience, which is inefficient and highly subjective; (4) Lack of standardized high-resolution datasets: There is a lack of publicly available patterned wafer bright-field image resources, which limits the research on anomaly detection based on deep learning.

[0006] Therefore, there is an urgent need for a bright-field acquisition system and supporting data construction method that balances high imaging quality, automated acquisition, and data reliability to support research on intelligent defect detection of wafer microstructures. Summary of the Invention

[0007] To address the problems of uneven illumination, insufficient depth of focus, and low annotation efficiency in existing bright-field detection methods, this invention proposes a patterned wafer microstructure data acquisition system and dataset construction method under bright-field conditions. This system can obtain high-depth-of-field, high-contrast microscopic images and construct a high-quality anomaly detection dataset that can be directly used for deep learning model training.

[0008] This invention adopts the following technical solution: a method for constructing a patterned wafer microstructure dataset under bright field conditions, comprising the following steps:

[0009] Step 1, Sample Fixation and Initialization: Fix the patterned wafer sample to be inspected on the stage of the three-axis electric displacement platform, and set the initial scanning parameters according to the wafer size and requirements;

[0010] Step 2, Illumination Adjustment: Set the illumination mode and adjust the brightness through the adjustable light source component to obtain uniform bright field illumination and ensure that the sample surface and the optical axis remain horizontal.

[0011] Step 3, Displacement Platform Control and Scanning: Control the three-axis electric displacement platform to perform X / Y plane scanning according to the predetermined trajectory, trigger the imaging process at each acquisition point, and perform full coverage acquisition of the sample surface image;

[0012] Step 4, Autofocus and Multi-Focal Plane Acquisition: At each acquisition point, the microscope objective performs autofocus scanning along the Z-axis to acquire a sequence of multi-focal plane images, and fuses the acquired multiple focal plane images to generate the optimal focal plane image.

[0013] Step 5, Image Cropping and Preprocessing: Use image processing software to crop the optimal focal plane image, and perform noise suppression, illumination equalization, background smoothing and distortion correction operations;

[0014] Step 6, Image Classification and Sample Screening: Classify the preprocessed image data to distinguish between normal and abnormal samples; the abnormal samples enter the subsequent annotation stage, while the normal samples are used to construct defect-free control data for training the anomaly detection algorithm.

[0015] Step 7, Anomaly Sample Labeling and Mask Generation: Anomaly areas are labeled at the pixel level through an interactive labeling interface to generate defect masks, and then classified and recorded according to defect type;

[0016] Step 8: Dataset Generation and Export: Integrate image data and defect masks to generate a structured dataset for algorithm research. Store the dataset using a unified naming convention and hierarchical directory structure, and support exporting it to a general machine learning format.

[0017] As a preferred embodiment, the initial scanning parameters include: scanning range, step spacing, and focal length scanning range; the three-axis electric displacement platform is driven by a stepper motor to perform matrix or linear scanning according to the preset path and scanning parameters of the control terminal.

[0018] As a preferred embodiment, the adjustable light source assembly includes a ring-shaped LED bright field light source that supports continuously adjustable brightness. Its illumination parameters are adjusted according to the real-time preview results of computer software to suppress specular reflection and enhance the contrast of microstructure textures.

[0019] As a preferred embodiment, step 4 involves fusing the obtained multiple focal plane images to generate the optimal focal plane image, as follows:

[0020] Step 4.1: Calculate the image sharpness index for a local window of each focal plane image;

[0021] Step 4.2: Determine the weight of the local window based on the aforementioned sharpness index;

[0022] Step 4.3: Weave the focal planes according to their weights to obtain a final image with enhanced depth of field and clear details.

[0023] As a preferred embodiment, the region cropping in step 5 involves cropping the original bright-field micrograph according to a preset region extraction rule, retaining the effective imaging region containing patterned structures, and removing edge regions, unpatterned regions, and regions with uneven illumination or out-of-focus blur, to form image blocks of uniform size.

[0024] The cropped image is sequentially subjected to noise suppression, illumination equalization, background smoothing, and geometric distortion correction to eliminate system imaging noise, improve brightness uniformity, enhance texture stability, and compensate for spatial distortion caused by the optical system, thus obtaining preprocessed image data.

[0025] As a preferred embodiment, step 7 involves pixel-level annotation of the abnormal regions. The annotation content includes: the category label of the abnormal region, the pixel-level contour point set of the abnormal region, the annotation shape type, the group identifier, and additional flags; and the path and size metadata of the original image are recorded. The annotation results are stored in JSON format, with each abnormal region recorded as an independent entry, and multiple defect regions can be annotated simultaneously.

[0026] As a preferred embodiment, the method for generating the defect mask in step 7 is as follows:

[0027] Step 7.1: Read the annotation JSON file, parse it to obtain the shape type and pixel-level contour point set of each annotation item, and determine the corresponding region of each annotation item in the image;

[0028] Step 7.2: On a blank mask of the same size as the original image, fill each labeled area to generate a mask that only highlights the labeled areas, and save it in .png format.

[0029] The present invention also provides a data acquisition system for patterned wafer microstructures in a bright field environment, for implementing the above-mentioned dataset construction method, including: an adjustable light source component, a microscope imaging component, and a three-axis electric displacement platform.

[0030] The adjustable light source assembly, positioned above the detection platform, includes a ring-shaped LED bright-field light source. It supports continuously adjustable brightness and is used to adjust the illumination distribution under different sample reflectivity characteristics, improving illumination uniformity and reducing the effects of specular reflection. Light source control is typically set manually before acquisition to ensure good brightness and contrast in the acquired image.

[0031] The microscope imaging assembly, working in conjunction with a three-axis motorized displacement platform, includes a microscope objective, an industrial camera, and an image acquisition module, used to acquire images of the microstructure of the wafer surface. The microscope objective is connected to the industrial camera and moves gradually within a set Z-axis direction to achieve multi-focal plane imaging. The autofocus and image fusion unit of the image acquisition module automatically selects or fuses multi-focal plane images according to the image sharpness index to obtain a final image with enhanced depth of field and clear details.

[0032] The three-axis electric displacement platform, driven by a stepper motor, is used for precise positioning and motion control in the X, Y plane and Z axis directions. It can realize sample scanning and point-by-point acquisition along a predetermined trajectory. The motion control of the platform is synchronized with the imaging acquisition process, ensuring focal stability and spatial consistency, and improving acquisition efficiency and repeatability.

[0033] As a preferred embodiment, the microscope objective lens is connected to the industrial camera via a standard C interface to ensure optical axis concentricity and imaging stability; the industrial camera communicates with the host computer via a USB interface for real-time image acquisition, display, and storage.

[0034] As a preferred embodiment, the stepper motor has a resolution of 0.1μm and a maximum stroke range of 100mm in both the X and Y directions and 25mm in the Z direction.

[0035] Compared with the prior art, the present invention, employing the above technical solution, has the following technical effects:

[0036] 1. Compared to traditional single-focal-plane acquisition methods or fully automated inspection processes, the method of this invention effectively improves image clarity and enhances the identifiability of subtle defects while maintaining acquisition efficiency. Furthermore, this invention effectively balances the degree of automation in the acquisition process with the accuracy of data annotation through a labeling mechanism that combines automated acquisition with manually controllable annotation, thereby improving data consistency and reliability.

[0037] 2. The method of the present invention can solve the problems of uneven illumination and insufficient depth of focus in existing bright field detection by multi-focal plane scanning and automatic fusion, obtain microscopic images with high depth of field and high contrast, and construct a high-quality anomaly detection dataset that can be directly used for training deep learning models.

[0038] 3. The system of this invention is suitable for patterned wafer microstructure acquisition in scientific research and industrial inspection scenarios. Especially under conditions of large sample differences and unstable automatic algorithms, it can realize a high-fidelity and high-consistency data construction process and can be widely used in wafer manufacturing inspection, microstructure characterization, intelligent visual inspection and deep learning anomaly detection model training and other scenarios. Attached Figure Description

[0039] Figure 1 The images show the overall framework, front view, and right view of the patterned wafer microstructure data acquisition system of this invention.

[0040] Figure 2 This is a flowchart of the method for constructing a patterned wafer microstructure dataset according to the present invention;

[0041] Figure 3 This is a real photograph of a wafer placed on a three-axis electric displacement platform according to an embodiment of the present invention.

[0042] Figure 4 This is an example diagram of the hierarchical structure of the dataset in an embodiment of the present invention. Detailed Implementation

[0043] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of the application will be further described in detail below with reference to the accompanying drawings. The described embodiments are only a part of the embodiments involved in this invention. All non-innovative embodiments based on these embodiments by other researchers in the art are within the protection scope of this invention. Furthermore, the step numbers in the embodiments of this invention are only set for ease of explanation and do not limit the order of the steps. The execution order of each step in the embodiments can be adaptively adjusted according to the understanding of those skilled in the art.

[0044] Example 1

[0045] A data acquisition system for patterning wafer microstructures in a bright-field environment is provided, and its overall architecture, front view, and right view are shown below. Figure 1 As shown, it mainly consists of three components: an adjustable light source component, a microscopic imaging component, and a three-axis electric displacement platform. The components work together through hardware connection and software control to achieve automated acquisition of high-resolution images of the wafer surface.

[0046] Specifically, the adjustable light source assembly is located above the microscope objective and uses a ring-shaped LED light source structure to provide uniform bright-field illumination.

[0047] Before the formal data collection, the operator adjusts the brightness based on the real-time imaging preview results on the computer software to ensure that the target area has uniform brightness and no obvious reflections or dark areas.

[0048] During use, the operator can manually adjust the light intensity using the brightness control potentiometer (output current adjustable from 0 to 100%) to adapt to the reflection characteristics of different sample surfaces.

[0049] After adjustment, the light source is fixed and remains unchanged during the acquisition process to ensure stable and consistent illumination throughout the scanning process.

[0050] Specifically, the microscopic imaging components include: a high-resolution industrial camera (AIC-502C-USB) and a microscope objective (in this embodiment, a domestically produced infinity metallurgical objective is used).

[0051] In this embodiment, the industrial camera has a resolution of 2592×1944 pixels, supports 32-bit RGB image output, and has a pixel magnification of approximately 0.29×. The microscope objective lens magnification can be selected from 20× (IOMPSO138420X-VIS) and 50× (IOMPSO140450X-VIS) (20× was selected in this embodiment), which can achieve bright-field imaging with sub-micron resolution.

[0052] The microscope objective lens and the industrial camera are connected via a standard C interface to ensure optical axis concentricity and imaging stability; the industrial camera communicates with the host computer via a USB interface to achieve real-time image acquisition, display and storage.

[0053] Specifically, the three-axis electric displacement platform is driven by a high-precision stepper motor with a resolution of up to 0.1μm and a maximum travel range of 100mm in the X and Y directions and 25mm in the Z direction.

[0054] During the data acquisition process, the three-axis electric displacement platform performs matrix or linear scanning according to the preset path on the control terminal. The specific sampling step size (e.g., 50μm, 100μm) and coverage area can be set according to the sample size and requirements.

[0055] At each scanning point, the system automatically drives the Z-axis to scan in steps within a range of ±20μm, acquiring multiple focal plane images, which are then fused to obtain a final image with enhanced depth of field and clear details.

[0056] Example 2

[0057] A method for constructing a patterned wafer microstructure dataset under bright field conditions based on the system of Embodiment 1 is provided, the process of which is as follows: Figure 2 As shown, by comprehensively utilizing automated hardware control, adjustable illumination imaging, and manual annotation mechanisms, the aim is to achieve high-precision and high-consistency wafer surface image acquisition and defect data construction.

[0058] Step 1: Sample fixation and initialization.

[0059] like Figure 3 As shown, the patterned wafer sample to be inspected is stably fixed on the stage of a three-axis electric displacement platform, ensuring that the sample surface remains horizontally aligned with the optical axis. Initial parameters such as scanning range, step spacing, and focal length scanning range are set according to the wafer size and requirements.

[0060] In this embodiment, the industrial camera has a resolution of 2592×1944, an imaging magnification of approximately 0.29×, and outputs a 32-bit RGB image format to ensure sufficient spatial resolution and color information fidelity.

[0061] Step 2, Light Adjustment.

[0062] The lighting mode can be set by adjusting the light source component, and the brightness can be adjusted to obtain uniform bright field illumination in the imaging area.

[0063] The lighting parameters can be manually adjusted by the operator based on the real-time preview results on the computer software to suppress specular reflection and enhance the contrast of microstructure textures. Figure 3 The yellow adjustable knob in the system is used to control the light intensity, thereby achieving optimal imaging brightness for different materials and structural surfaces.

[0064] Step 3: Displacement platform control and scanning.

[0065] The three-axis electric displacement platform is controlled to perform X / Y plane scanning according to a predetermined trajectory, triggering the imaging process at each acquisition point to achieve full coverage acquisition of the sample surface.

[0066] Step 4: Autofocus and multi-focal plane acquisition.

[0067] At each acquisition point, the microscope objective performs autofocus scanning along the Z-axis, acquiring a sequence of multi-focal-plane images. The system then fuses these multiple focal-plane images to generate a final image with enhanced depth of field and clear structure. The specific method is as follows:

[0068] (a) Calculate the image sharpness index for a local window of each focal plane image;

[0069] (b) Determine the weight of the local window based on the aforementioned sharpness index;

[0070] (c) The focal planes are weighted and fused to obtain a final image with enhanced depth of field and clear details, which is taken as the optimal focal plane image.

[0071] Step 5: Image cropping and preprocessing.

[0072] The optimal focal plane image is cropped using image processing software. In this embodiment, the image processing software is GIMP 3.0.2-1. The original brightfield micrograph is cropped according to a preset region extraction rule, retaining the effective imaging area containing patterned structures and removing edge areas, unpatterned areas, and areas with uneven illumination or out-of-focus blur, thereby forming image blocks of uniform size. In this embodiment, all image data is cropped into 1024×1024 pixel image blocks.

[0073] Subsequently, noise suppression, illumination equalization, background smoothing, and geometric distortion correction are performed on the cropped image in sequence to eliminate system imaging noise, improve brightness uniformity, enhance texture stability, and compensate for spatial distortion caused by the optical system.

[0074] Step 6: Image classification and sample selection.

[0075] Images are manually inspected and categorized to distinguish between normal and abnormal samples. Abnormal samples will proceed to the subsequent annotation stage. Normal samples, on the other hand, are used to construct defect-free control data to support the training of subsequent anomaly detection algorithms.

[0076] Step 7: Anomaly sample labeling and mask generation.

[0077] Manual annotation is performed on abnormal areas at the pixel level through an interactive annotation interface to generate defect masks, and then the defects are classified and recorded according to their types.

[0078] In this embodiment, the annotation tool used is labelme.

[0079] The annotation content includes: category label of the abnormal region, pixel-level contour point set of the abnormal region, annotation shape type, group identifier and additional flags; and records metadata such as path and size of the original image.

[0080] Annotation results: stored in JSON format, where each abnormal area is recorded as an independent entry, and multiple defect areas can be annotated simultaneously.

[0081] The method for generating defect masks is as follows:

[0082] First, the annotation JSON file is read and parsed to obtain the shape type and pixel-level contour point set of each annotation item, thereby determining the corresponding region of each annotation item in the image.

[0083] Then, on a blank mask of the same size as the original image, each labeled area is filled to generate a mask that only highlights the labeled areas. Finally, the mask is saved in .png format.

[0084] Step 8: Dataset generation and export.

[0085] Image data and defect masks are integrated to generate a structured dataset. The dataset is stored using a unified naming convention and a hierarchical directory structure, and supports export to a common machine learning format.

[0086] In this example, the export format is a combination of .png and .json, and the hierarchical structure of the dataset can be selected. Figure 4 The structure.

[0087] Compared to traditional single-focal-plane acquisition methods or fully automated inspection processes, the method in this embodiment can effectively improve image clarity and enhance the identifiability of subtle defects while maintaining acquisition efficiency. Furthermore, by combining automated acquisition with manually controllable annotation, a balance is effectively struck between the degree of automation in the acquisition process and the accuracy of data annotation, thereby improving data consistency and reliability.

[0088] The above description is only a preferred embodiment of the present invention. It should be noted that for those skilled in the art, several improvements and modifications can be made without departing from the principle of the present invention, and these improvements and modifications should also be considered within the scope of protection of the present invention.

Claims

1. A method for constructing a patterned wafer microstructure dataset under bright-field conditions, characterized in that, Includes the following steps: Step 1, Sample Fixation and Initialization: Fix the patterned wafer sample to be inspected on the stage of the three-axis electric displacement platform, and set the initial scanning parameters according to the wafer size and requirements; Step 2, Illumination Adjustment: Set the illumination mode and adjust the brightness through the adjustable light source component to obtain uniform bright field illumination and ensure that the sample surface and the optical axis remain horizontal. Step 3, Displacement Platform Control and Scanning: Control the three-axis electric displacement platform to perform X / Y plane scanning according to the predetermined trajectory, trigger the imaging process at each acquisition point, and perform full coverage acquisition of the sample surface image; Step 4, Autofocus and Multi-Focal Plane Acquisition: At each acquisition point, the microscope objective performs autofocus scanning along the Z-axis to acquire a sequence of multi-focal plane images, and fuses the acquired multiple focal plane images to generate the optimal focal plane image. Step 5, Image Cropping and Preprocessing: Use image processing software to crop the optimal focal plane image, and perform noise suppression, illumination equalization, background smoothing and distortion correction operations; Step 6, Image Classification and Sample Screening: Classify the preprocessed image data to distinguish between normal and abnormal samples; the abnormal samples enter the subsequent annotation stage, while the normal samples are used to construct defect-free control data for training the anomaly detection algorithm. Step 7, Anomaly Sample Labeling and Mask Generation: Anomaly areas are labeled at the pixel level through an interactive labeling interface to generate defect masks, and then classified and recorded according to defect type; Step 8: Dataset Generation and Export: Integrate image data and defect masks to generate a structured dataset for algorithm research. Store the dataset using a unified naming convention and hierarchical directory structure, and support exporting it to a general machine learning format.

2. The method for constructing a patterned wafer microstructure dataset under bright-field environment according to claim 1, characterized in that, The initial scanning parameters include: scanning range, step spacing, and focal length scanning range; the three-axis electric displacement platform is driven by a stepper motor to perform matrix or linear scanning according to the preset path and scanning parameters of the control terminal.

3. The method for constructing a patterned wafer microstructure dataset under bright-field environment according to claim 1, characterized in that, The adjustable light source assembly includes a ring-shaped LED bright field light source that supports continuously adjustable brightness. Its illumination parameters are adjusted according to the real-time preview results of computer software to suppress specular reflection and enhance the contrast of microstructure textures.

4. The method for constructing a patterned wafer microstructure dataset under bright-field environment according to claim 1, characterized in that, Step 4 involves fusing the obtained multiple focal plane images to generate the optimal focal plane image. The method is as follows: Step 4.1: Calculate the image sharpness index for a local window of each focal plane image; Step 4.2: Determine the weight of the local window based on the aforementioned sharpness index; Step 4.3: Weave the focal planes according to their weights to obtain a final image with enhanced depth of field and clear details.

5. The method for constructing a patterned wafer microstructure dataset under bright-field environment according to claim 1, characterized in that, Step 5, the region cropping, involves cropping the original bright-field micrograph according to a preset region extraction rule, retaining the effective imaging region containing patterned structures, and removing edge regions, unpatterned regions, and regions with uneven illumination or out-of-focus blur, to form image blocks of uniform size. The cropped image is sequentially subjected to noise suppression, illumination equalization, background smoothing, and geometric distortion correction to eliminate system imaging noise, improve brightness uniformity, enhance texture stability, and compensate for spatial distortion caused by the optical system, thus obtaining preprocessed image data.

6. The method for constructing a patterned wafer microstructure dataset under bright-field environment according to claim 1, characterized in that, Step 7 describes pixel-level annotation of abnormal regions. The annotation content includes: category label of abnormal region, pixel-level contour point set of abnormal region, annotation shape type, group identifier and additional flag bits, and records the path and size metadata of the original image. The annotation results are stored in JSON format, with each abnormal region recorded as an independent entry, and multiple defect regions can be annotated at the same time.

7. The method for constructing a patterned wafer microstructure dataset under bright-field environment according to claim 6, characterized in that, Step 7, generating the defect mask, is performed as follows: Step 7.1: Read the annotation JSON file, parse it to obtain the shape type and pixel-level contour point set of each annotation item, and determine the corresponding region of each annotation item in the image; Step 7.2: Fill each labeled area on a blank mask of the same size as the original image to generate a mask that only highlights the labeled areas, and save it in .png format.

8. A data acquisition system for patterned wafer microstructures under bright-field conditions, used to implement the dataset construction method according to any one of claims 1 to 7, characterized in that, Includes: an adjustable light source assembly, a microscope imaging assembly, and a three-axis motorized displacement platform; The adjustable light source assembly, arranged above the detection platform, includes a ring-shaped LED bright field light source that supports continuously adjustable brightness. Based on the different reflectivity of the samples, the light distribution is adjusted before data acquisition to provide uniform bright field illumination. The microscope imaging assembly, working in conjunction with a three-axis motorized displacement platform, includes a microscope objective, an industrial camera, and an image acquisition module, used to acquire images of the microstructure of a wafer surface; the microscope objective is connected to the industrial camera and moves gradually within a set Z-axis direction to achieve multi-focal plane imaging; the autofocus and image fusion unit of the image acquisition module automatically selects or fuses multi-focal plane images according to image sharpness indicators to obtain the final image; The three-axis electric displacement platform is driven by a stepper motor and performs precise positioning and motion control in the X, Y planes and Z axis directions to ensure focal stability and spatial consistency, thereby realizing sample scanning and point-by-point acquisition along a predetermined trajectory.

9. The patterned wafer microstructure data acquisition system under bright field environment according to claim 8, characterized in that, The microscope objective lens is connected to the industrial camera via a standard C interface to ensure optical axis concentricity and imaging stability; the industrial camera communicates with the host computer via a USB interface for real-time image acquisition, display and storage.

10. The patterned wafer microstructure data acquisition system under bright-field environment according to claim 8, characterized in that, The stepper motor has a resolution of 0.1μm and a maximum stroke range of 100mm in both the X and Y directions and 25mm in the Z direction.

Images