The Xiris Blog

How to Get the Best View of an Open Arc Weld

Posted by Cameron Serles on Thursday, July 17, 2014 @ 06:00 PM

Attaining a good image of a weld and the surrounding background has been a struggle ever since video cameras for welding became available.  The problem has always been the range of brightness that occurs during welding: the ratio between the maximum and minimum light intensity is usually too great for a standard camera to measure properly.  Standard cameras on the market today can typically measure about 1,000 levels of brightness between the maximum and minimum light levels in an image.  However, in a typical open arc welding environment, there is a brightness range that can exceed 10,000,000 levels of brightness between the brightest portion of the welding arc, and the darker areas surrounding the weld.  Using a standard camera to image such a weld will create an image similar to the image below on the left, where the camera sensor will image the scene up to a point and then saturate when it gets too bright. This causes the bright areas of the image to appear as a white blur.


To solve this problem, Xiris Automation has developed the XVC-O View Camera that uses advanced electronics with logarithmic sensitivity to be able to see more than 10,000,000 levels of brightness in an image.  As a result, more image detail is visible than ever seen before. The detail of the weld arc, the shielding gas, weld pool, torch tip, and weld seam can all clearly be seen.  The image below on the right is an image taken from the XVC-O camera of an open arc welding process. The weld arc is no longer saturated and is clearly visible as is the detail of the background, providing better quality information for the weld operator.


GOOOOOOD resized 600       Standard Camera Image of a Weld                      Xiris XVC-O Camera Image of a Weld

With the ability to see more detail of the weld arc and the surrounding environment, welding technicians are able to use the XVC-O to better control their welding processes through better quality assurance and process feedback. 

To see examples of the video quality possible with the XVC-O across a variety of welding processes and materials, please see our Weld Video Library here.


Topics: weld camera, weld inspection, Laser welding, welding automation, weld environment, Machine Vision, image processing, Education, Welding Process, weld video, Xiris, image contrast

Better Images, Better Instruction, Better Welding Students!

Posted by Cameron Serles on Wednesday, February 12, 2014 @ 03:50 PM

Training a new group of welding students can have a number of challenges for even the best instructors: getting all the students around the weld head to be able to see what is going on; a limited number of hours the instructor has available for actually performing the welding; how to see all the features of the weld arc as well as the background information, and how to make sure that all students are marked fairly and objectively. 

When educating welding students, providing them with the ability to view the detail of the weld tip as well as the environment around the weld tip (such as the weld seam and weld pool) is important for them to learn all the parameters of the welding process.  To overcome the visual monitoring challenges created by the presence of a very bright light source (the weld arc), as well as dark areas in the image (the background around the weld tip), a camera with a wide dynamic range of imaging is required.  Reliable visualization of the environment around the weld tip is necessary to control and adjust the welding process found on most modern welding processes.  In addition, the ability to record video and play it back to the students can provide multiple benefits for teaching and correcting welding techniques.

 Blog 141212 students resized 600

Image courtesy of Casper College

They Can’t All See the Details…. 

New developments in electronics has led to the creation of a new type of camera that is able to accommodate the full range of light present at a weld head during welding, allowing welding to be taught in a way it has never been taught before!

By providing a good quality image of the weld tip and background, welding instructors and their students can remotely monitor a weld demonstration and record the results for off-line feedback.  By using a camera to view the weld demonstration, the students can verify that the tip is in position and that all the welding inputs (welding wire, shielding gas, etc.) are being properly fed.  Because the area around the weld demonstration is typically quite congested for class sizes more than a few students, using a camera mounted at the welding tip allows the students to clearly view the welding process remotely.  The video can also be replayed back, off-line in the classroom for instruction, marking or review purposes. 

 Blog 141212 xiris resized 600

The Solution: a Xiris XVC-O View Camera for Teaching Welding


Using a View Cameras in the classroom to teach welding results in:

  • —  A more Enjoyable Learning Experience for the Students
  • —  Less Time Required to Achieve Results
  • —  Reduced Material Consumption
  • —  A Video Library of Standard Applications for Review / Consulting / Analysis
  • —  Easier to Explain New Welding Techniques
  • —  Better Support for Students’ Technical Projects
  • —  Research Tool

Join the growing number of Welding Educational Institutions who have added a Xiris XVC-O View Camera to their classrooms. Improve welding instruction and achieve the numerous benefits!

To read educator's personal testimonials below



For more information on how Xiris Weld Cameras can augment your welding education program, please visit 

Sign up to receive our Weld Video of the Month

Topics: remote monitoring, weld camera, weld inspection, Laser welding, Machine Vision, image processing, field of view, welding instruction, Education, High Dynamic Range, laser-based monitoring, image contrast

The Technology Behind Every Xiris System

Posted by Cameron Serles on Tuesday, January 21, 2014 @ 03:42 PM

Every machine vision system developed by Xiris is based upon an internal image processing library developed by Xiris over a 20 year period including thousands of hours of coding and testing.  The library has been built up to include a number of key algorithms to perform specific imaging tasks, and includes a number of tools created for maximum specification.

 Pattern Match Tool

Figure 1: This image demonstrates the capabilities of the Pattern match tool

The Edge Tools are used to very precisely locate single edges or edge pairs in a straight line or along an arc.  Best used for precisely gauging the distance between two edges or object location (finding one or two edges precisely to locate a corner, or feature), the software has been scientifically proven in a lab environment to be accurate to better than 1/20th of a pixel.

The Blob Tool is used to perform shape analysis of randomly oriented objects in an image with over 60 different measured features.  It can be used to determine if object meets specific criteria with individual thresholds available for each parameter to quickly select the features of interest, including: Area, Perimeter, Equivalent Diameter, Centroid, Orientation, Second Moments, Bounding Box, Circularity, Eccentricity and many others.

The Pattern Match (or Search) Tool, as shown above, is used to perform pattern matching by locating two dimensional objects in a very accurate manner (1/4 pixel or better).  The Pattern Match Tool is very useful when trying to find an object in a complex scene. Additionally, this tool can be used to determine how well an object matches its ideal, or golden part. 

Used to identify lighting variations, the Light Meter (or Histogram) Tool monitors overall intensity changes or simple part presence. Another important application of this tool is the modification of the Brightness or Contrast of the overall image.

Print Inspection uses a golden template pattern, built over a series of taught images, to compare with the image under inspection.  The Euclidean distance between each overlaid pixel is analyzed to determine if the image under inspection meets with predefined defect criteria.  During the teaching phase, the tool automatically determines appropriate inspection thresholds for each pixel.

Color Tools can be used to verify or recognize a region of color on a component by comparing the candidate color to a series of pre-taught, or known colors.  The color processing can be done inRGB (Red, Green, Blue), HLS (Hue, Luminance, Saturation), CMYK (Cyan, Magenta, Yellow, Black), XYZ or other color spaces.  These tools are often used for product identification, print image quality, feature analysis.

Symbology Tools help read or verify various types of symbols, including multi-format 1-D and two-D bar codes, as well as Optical Character Verification and Recognition.  Characters can be verified or recognized down to 24x24 pixels in size.

Other tools in our software package include Temporal Tools, which can track position over time, a number of Surface Inspection tools, which range from feature detection to defect classification, and 3D Imaging with the use of Laser Triangulation, used to extract the 3-dimensional shape of a surface

Our other Image Processing Tools are used to enhance an image for the benefit of an operator, including morphology (shape based processing such as erosion/dilation or opening/closing), and neighborhood processing using convolutions, including: Sobel, Averaging, Sharpening, Low Pass, Median, Watershed, and others.

All of these tools were created to help the operator make better decisions based upon what they are able to monitor. By using its own software imaging library in its machine vision systems, Xiris is able to provide custom algorithms to suit some very specific market requirements, achieving greater speed and performance benefits that would be otherwise unavailable were a general purpose imaging library to be used. The customizable nature of this software toolkit makes a Xiris system essential for a variety of applications. 


Have an application that could benefit from these tools? Contact us, we always welcome the opportunity to discuss new initiatives! 


Topics: remote monitoring, camera selection, quality control, weld camera, weld camera, weld inspection, Machine Vision, image processing, field of view, High Dynamic Range, laser-based monitoring, image contrast

How Image Contrast Affects Remote Weld Monitoring

Posted by Cameron Serles on Tuesday, April 23, 2013 @ 12:53 PM

Image contrast of a weld monitoring scene is a measure of the difference in brightness between the light in the scene (i.e., the weld arc area) and the dark areas in the scene (i.e., the background area of the weld).

To fully understand image contrast, it’s important to discuss the image histogram—a chart of all the brightness values and the number of pixels that are at each level of brightness. (Typically, we display images with 8-bit detail, which corresponds to 256 levels of brightness.)

Broad histograms indicate a weld environment scene with significant contrast, whereas narrow histograms signify less contrast in the scene, which can cause the resulting image to appear flat or dull.

The contrast variance can be caused by any combination of subject matter and lighting conditions. For example, images of a weld process taken with lots of smoke present will have low contrast, while those taken under a clear, bright arc will have higher contrast. Likewise, weld pools with lots of texture where light is not evenly reflected provide higher contrast than flat polished sections of metal that reflect at a near-constant brightness.

Image technology also affects the contrast in the image of the scene. If the brightest pixels aren’t saturated, then the resulting image will usually have a higher contrast.  Typically, better quality images can be obtained from a High Dynamic Range imaging system than a regular camera because it can provide images with greater contrast.

Why You Want High Contrast

High contrast can have a significant positive visual impact on a weld image by enhancing the texture of the details in the image, as shown by comparing the high-contrast and low-contrast images below. The high-contrast image has deeper shadows and more pronounced highlights, creating texture in the image that "pops" out at the viewer.


High-contrast image from Weld Camera with High Dynamic Range imaging

 High-Contrast Image



Low-contrast image from Weld Camera with High Dynamic Range imaging

Low-Contrast Image

This enhanced texture allows operators to more-clearly see the precise details necessary to make in-process adjustments—leading to greater accuracy, faster decisions, and less strain on the eyes. In TIG welding, for example, when the operator wants to check for surface contamination, a higher contrast image will allow for more features at the edge of the weld pool to be visible, such as dross, bubbles or bead undercut that can indicate the quality of the weld.

In some kinds of welding, such as Plasma welding, a higher contrast image can show features that would not be visible otherwise, such as the seam that is visible through the weld (see image below). With such visibility, the operator can use the camera as an aid for seam guidance.


Weld Seam Through the Weld Cone as Shown in Image from Weld Camera 

Seam Through the Weld Cone


Differing Contrast in the Same Image

Contrast can also vary between different regions within the same image due to subject matter and lighting differences in the scene. To illustrate, we can partition an image of an MIG weld process into three separate regions—each with its own distinct histogram.


Variable Contrast Within MIG Weld Image

Variable Contrast Within MIG Weld Image


The upper region contains a fairly broad contrast, but only across a part of the histogram as it does not have any super bright features in it.

The middle region contains the most contrast of all three regions because the image includes the direct light generated by the welding arc rather than light that first reflects off the surface of any metal surface in the field of view. This higher contrast produces stronger edges in the region, as well as stronger highlights in the weld arc area.

The bottom region contains smoke that captures some of the diffuse, reflected light from the weld arc and thus has lower contrast—similar to images you’d get if you were taking photographs in the fog without a bright, direct light source.

The sum of the histograms in all three regions creates the overall histogram of the image, seen below.

Whole image histogram 

Whole Image Histogram



High-contrast scenes of welding processes contain more texture and more visible features than low-contrast scenes, resulting in improved operator perception and understanding of the weld process. By creating higher-contrast images, details in a weld process become visible that would not be seen otherwise.

Topics: weld environment, High Dynamic Range, image contrast

Latest Posts

Follow Me