All Products
-
Stencil Inspection Machine System
-
Squeegee Inspection Machine
-
Digital measuring guage
-
AOI Stencil Inspection Machine
-
Wafer Inspeciton Machine
-
Screen Printing Stencil Inspection Machine
-
Stencil/PCB/Fixture Cleaning Machine
-
Label printing and Inspection Machine
-
Reflow Oven Temperature Curve Tester
-
PCB Barcode Inspection Machine
-
FPC Stiffener Adhesive Machine
AI Automatic AOI Machine Silk Screen Mesh Stencil Inspection Machine for 736*736mm Metal Mesh testing
| Place of Origin | China |
|---|---|
| Brand Name | Sunvishe |
| Certification | CE, ROSH |
| Model Number | SV-M8 |
| Minimum Order Quantity | 1 |
| Packaging Details | Plywood case+vacuum package |
| Delivery Time | 4 weeks |
| Supply Ability | 30 units machine per month |
Product Details
| Application | New Stencil Incoming Inspection Or Inspection After Cleaning | Color | White |
|---|---|---|---|
| Stencil Size | 736mmx736mm | Warranty | 12 Months |
Product Description
AI Precision Automatic Optical
Inspection Machine for Silk
Screen Mesh Stencil
SV-M8 Series
Brief Introduction to Main Functions
This high-precision screen stencil inspection system is mainly used for screen appearance
defect inspection, say the pattern, photoresist thickness, burrs, layer deformation, and mesh integrity
etc. Defect detection is carried out automatically through optical imaging, blind
inspection algorithm, comparison algorithm and AI algorithm.
A. Automatically scan and identify CAM data, and automatically parse CAM design data.
B. The automatic optical system scans product images for data processing, extracting effective features
and coordinates.
C. Automatically identify and classify defects, display actual defect images for secondary confirmation;
indicate defect positions via laser positioning targets and output defect coordinate list.
D. The inspection workstation can automatically read the inspected product defect data and display
them one by one for fast and convenient repair.
E. Inspection information is saved in real time for later traceability, statistics and SPC analysis.
F. Full-process inspection is available for screens (blank screen, pre-pressed screen, finished PI screen,
finished emulsion screen, finished metal mask).
G. Thresholds can be set independently as needed to avoid over-quality control and facilitate unified
1. Overview
In traditional manufacturing plants such as precision screen production, the inspection of tiny defects is performed
manually. However, manual inspection of such small and highly variable defects has many problems, including high
labor intensity, high operational repeatability, high false detection and misjudgment rates, and product quality greatly
affected by human factors. In addition, each factory produces a wide variety of screens with different specifications,
which further increases the workload of screen defect inspection. Manual inspection cannot achieve stable, accurate
and high-efficiency screen defect detection, resulting in the failure of manufacturers to stably supply qualified
products and meet market demand, which will seriously reduce factory productivity. Manufacturing enterprises are
increasingly adopting more stable and accurate equipment in the production process to improve production efficiency.
As one of the important cutting-edge technologies in artificial intelligence, AI defect detection technology has
inherent advantages in screen defect inspection. AI defect detection technology makes screen defect inspection
non-contact, high-efficiency, cost-saving and productive, which is of great application value in detecting hidden
defects and preventing defective screen products. With the support of machine vision system, the inspection
efficiency of screen defect inspection in traditional manufacturing plants has achieved a qualitative leap, and
effectively solves the problems of difficult inspection standardization and high false detection rate.
The AI vision inspection system solution based on DLIA deep learning platform and AI industry models provided by
our company can efficiently complete algorithm matching, database and production line version upgrade for screen
defect categories. With the continuous iteration of screen defect types in production, the system achieves pixel-level
detection accuracy and millisecond-level inspection speed.
The SMI series high-precision AI screen optical inspection system, fully integrated with application scenarios,
realizes high-precision, high-efficiency and full-format defect detection for screens, which can significantly reduce
manual labor intensity, effectively improve the yield of screen printing screens, and enhance the market
competitiveness of enterprises. It has become a standard model for photovoltaic, microelectronic screens,
Brief Introduction to Main Functions
This high-precision screen stencil inspection system is mainly used for screen appearance
defect inspection, say the pattern, photoresist thickness, burrs, layer deformation, and mesh integrity
etc. Defect detection is carried out automatically through optical imaging, blind
inspection algorithm, comparison algorithm and AI algorithm.
A. Automatically scan and identify CAM data, and automatically parse CAM design data.
B. The automatic optical system scans product images for data processing, extracting effective features
and coordinates.
C. Automatically identify and classify defects, display actual defect images for secondary confirmation;
indicate defect positions via laser positioning targets and output defect coordinate list.
D. The inspection workstation can automatically read the inspected product defect data and display
them one by one for fast and convenient repair.
E. Inspection information is saved in real time for later traceability, statistics and SPC analysis.
F. Full-process inspection is available for screens (blank screen, pre-pressed screen, finished PI screen,
finished emulsion screen, finished metal mask).
G. Thresholds can be set independently as needed to avoid over-quality control and facilitate unifiedsemiconductor BGA ball mounting screens and PCB fields. With fixtures installed, it can also inspect sheets such as
nickel sheets. At present, all inspection functions have been effectively implemented and improved.
Physical Product - High-Precision Screen
1.1 Test Dimensions
Screen frame size:① 450×450 mm ② 356×356 mm ③ 736×736 mm
Screen inspection size: ① 210×210 mm ② 182×182 mm ③ 166×166 mm, etc.
Note: Inspection dimensions can be customized according to customer product requirements.
1.2 For existing screen inspection, image capture CT of a single product is less than 200S/PCS.
1.3 ① Machine overlook rate: Overlook rate of defects ≥5 μm: ≤1% (unknown defects not evaluated, software
continuously improved).
② False point false alarm rate: Probability of total false points exceeding 100 in a single inspection ≤15%.
3.Introduction to Equipment Core Components
| Core Component | Parameter Settings | QTY | Remarks |
| Main Unit | I7-12700K/32G×2/SSD 1TB/HDD 20TB, NVIDIA accelerated graphics card (HDMI×2 interface) | 1 | |
| Monitor | 27-inch HDMI high-definition monitor | 2 | |
| Camera & Lens |
Inspection machine: 20-mega pixel high-resolution color camera, 0.7 × telecentric low-distortion lens; 2-mega pixel high-definition video microscope, auto-focus lens |
1 | |
| Light Source | High-collimation backlight, indicator laser | 1 | |
| System | Wins 11 | 1 | |
| Core Software | Customized software + deep learning inspection algorithm | 1 | |
| Laser Displacement Sensor | Measuring range: ±15 mm, repeatability 10 μm, sampling frequency: 500/1000/2000/3000 Hz optional | 1 |
4.Specification
| Item | SV-M8 |
| Inspection Purpose | Screen defects |
| Inspection Items | Groove, broken yarn, broken grid (unbroken line), contamination, residual mesh wire, line hole blockage, notch (burr), misalignment, residual adhesive, film warpage, sand hole (pinhole), pattern abnormality, incomplete development, debonding, bubble, poor lamination, laser residue. Currently detectable: hole blockage, broken grid (unbroken line), foreign matter, dirt, sand hole (pinhole), crimp, broken yarn, residual mesh wire |
| Maximum Inspection Size | Standard machine: 736×736 mm (maximum screen frame size), customizable according to actual size |
| Equipment Accuracy | Standard machine: 620×620 mm (maximum effective inspection area); Pixel resolution: 2.36 μm, customizable according to actual needs |
| Positioning Accuracy | ±3 μm |
| Inspection Efficiency (sec/surface) | Estimated 100–200S/PCS (under the conditions of scanning graphic size 210 mm×210 mm and pixel resolution 2.36 μm, varies with the number of product defects) |
| Machine Overlook Rate | Overlook rate of defects ≥5 μm or 10 μm: ≤1% (unknown defects not evaluated, software continuously improved) |
| False Point False Alarm Rate | False point misjudgment rate: Probability of total false points exceeding 100 in a single inspection ≤20% |
| CAM Data Format | Gerber274X |
| Inspection Principle | CAM data comparison + AI deep learning + traditional algorithm |
| Light Source System | High-brightness LED light source (service life: more than 10,000 hours) |
| Camera | High-resolution industrial camera |
| Code Scanner | Wireless laser line positioning scanner, 1D code, 2D code, screen scannable |
| Operating System | Wins 11 |
| Dimensions | Inspection (L×W×H): 1200×1290×1520 mm (excluding three-color light tower) |
| Weight | 650 kg |
| Power Requirement | Single-phase AC220V, 50Hz, 2.5KW |
| MES | Support data upload, open and customize data interface, data structure and communication mode according to MES requirements |
| User Authority | User authority levels can be defined according to customer requirements |
Recommended Products