In the field of industrial visual inspection, the Keyence XG-8000 series and XG-7000 series are renowned for their ultra-high speed, multi camera support, and flexible processing capabilities. With the firmware upgrade to Ver.5.1 (XG-8000) and Ver.4.2 (XG-7000), the system has added a large number of practical features, including 1D barcode reading unit, image stitching unit, PROFINET industrial Ethernet communication, enhanced calculation functions, additional data output for target classification, FTP image transmission, and German interface support. The addition of these functions enables the XG series to not only complete traditional positioning, measurement, and appearance inspection, but also directly read barcodes, stitch large field of view images, and exchange data with PLCs at high speed through PROFINET, greatly expanding its application scope in intelligent manufacturing and automation production lines.
This article starts from the perspective of on-site application by engineers, systematically sorting out the configuration methods, typical application scenarios, common problems, and solutions of these new features, helping users quickly get started and avoid pitfalls.
1D barcode reading unit: allowing the visual system to directly read barcodes
1.1 Function Overview
Before Ver. 5.1/4.2, the XG series mainly relied on 2D code reading units to process QR codes, etc. The newly added 1D code reading unit supports the following code systems:
CODE39 (does not support Full ASCII conversion)
CODE128 (does not support FNC2 linking characters)
EAN/JAN/UPC (including EAN-8/13, UPC-A/E, and 2-digit/5-digit additional codes)
This unit can output rich result data such as barcode reading string, center position (X/Y coordinates), deflection angle, minimum bar width (pixels/module), reading stability (level 5), reasons for reading errors, etc. It also supports comparison with preset reference strings to achieve dynamic date, time, and serial number verification.
1.2 Key parameter settings and optimization
Suggested values for parameter description
Barcode type CODE39/CODE128/EAN/UPC can be selected as needed
When the minimum bar width module width (in pixels) position correction is turned off, it should be ≥ 1.5 pixels; Color image+position correction ≥ 3.2 pixels
The sensitivity of contrast feature extraction is set to Normal by default, and low contrast can be adjusted to Low
Barcode color black background white stripe/white background black stripe/both take a long time, it is recommended to fix them
The expected barcode direction of the reference angle is horizontal at 0 °, which can adapt to rotation
The maximum time for decoding timeout and single recognition is set to 3 seconds by default, which can be adjusted according to the beat
1.3 Advanced Features: Data Segmentation and Fixed Length Output
Data segmentation: The read string can be divided into up to 8 segments based on the starting position and length. Each segment can be output separately and used for reference comparison, suitable for composite information (such as batch number+production date+serial number).
Fixed length output: If the number of characters read is less than the set length, the specified characters (such as underscores) can be automatically filled in, making it convenient for PLC to receive fixed length data blocks.
Symbol identifier: Barcode type identification defined by JIS X 0530 can be added (such as "] J0" representing CODE39) to facilitate downstream system recognition of the code system.
1.4 Auto Tune
Click the "Auto Tune" button, and the system will automatically analyze the barcode in the current image, setting the most suitable contrast, minimum bar width, length range, etc. Possible reasons for automatic tuning failure:
There are multiple barcodes in the inspection area (unable to specify which one to read)
Excessive background noise or extremely low barcode contrast
The timeout period is too short
Solution: Reduce the inspection area to ensure that only one barcode is included; Increase lighting or use filters; Increase the decoding timeout value appropriately.
1.5 Error Handling and Common Problems
Possible causes and solutions for the phenomenon
Unable to read bar width too small (<1.2 pixels) to increase camera resolution or zoom in on objects
Read unstable and insufficient contrast. Enable the "contrast conversion" filter or use RGB grayscale extraction
Misunderstanding barcode with no static area around, increasing the detection area edge blank, or using a position correction unit
If the decoding time exceeds the set value, increase the decoding timeout, or optimize the image quality to reduce the computational load

Image stitching unit (XG-8700/8700L only)
2.1 Principles and Limitations
The image stitching unit can stitch 2-4 monochrome images (source images) into a large image with a maximum of 4864 × 4864 pixels according to the specified reference position and result position. The merged image can be used for all subsequent detection tools, such as pattern search, edge detection, etc.