IR Tech Tip: Identify Regions of Interest Before Analyzing Data

Infrared analysis software is commonly used to determine temperatures and profiles of objects under test. Temperature distributions including max/min/avg values, histograms and profiles are inevitably of particular interest. However, very often analyzing objects in an otherwise cluttered background can be problematic and result in skewed temperature distributions. In this IR Tech Tip, we’ll take a closer look at the power of the Altair Image Processing Software Suite. Designed to maximize the power of the Silver and Titanium high performance thermal imaging systems, Altair is rich with features and functions. However, unlike some thermal image processing programs, Altair is intuitive and user friendly.

Altair has many measurement tools that allow the user to measure the value of individual pixels, line profiles, or regions of interest. The Area, or ROI (region of interest) tool is probably the most used and most versatile measurement function within Altair.

Defining the ROI

Altair allows the user to define a region of interest (actually up to 25 ROIs) in one of three ways:

ROI-Rectangle

Any sized rectangle can be easily be drawn on a live or static image defining all the pixels within the ROI. In this example, a rectangle is defined over an integrated circuit on a Printed Circuit Board. Note that in the spreadsheet style data panel on the right, the maximum, minimum, average, and standard of deviation (along with other parameters) are displayed.

ROI-Polyline

Most real world objects, however are not perfectly rectangular. To correct for this, the user can create an area in the shape of a polygon with as many sides as possible. In this example, a multiple peaked target can be perfectly indentified.

ROI-Circle

For circular targets, the ROI Ellipse tool is perfect.

Histogram Distribution Within A Region of Interest

While many times it is simply valuable to view the max, min, and mean temperatures, many applications benefit from viewing the actual distribution of all temperatures within a region of intest. Using any shaped ROI, the user can have a real time histogram displayed adjacent to the image. The histogram plots the number of pixels verses their discrete temperatures.

In this image of two liquid materials mixing, the histogram on the right clearly shows the distribution of temperatures in the circle ROI. This histogram can be applied to a single frame or for each frame in a sequence.

ROI and Timing Graphs

One of the most useful uses of Altair is to track temperature (or digital values or radiance) over time. Altair incorporates a Timing Graph which was designed specifically to do just this. Operating either in a live mode or on a prerecorded sequence, the Timing Graph can plot values taken from any measurement tool, or tools on the image. Therefore if a user has multiple ROIs, line profiles or spots, the Timing Graph can plot these thermal values over time.

Using the same materials mixing sequence we diplayed previously, we can use the Timing Graph to show the trending of temperatures over time. On the right of the thermal image, a time versus temperature graph is created for the entire sequence. In this example, we have plotted the max, min, mean, and standard of deviation of the pixels within the ROI thoughout the complete sequence.

The vertical black line in the graph shows us which frame we are currently on; in this case frame 112 out of a total of 284 frames. (This can also be displayed in temporal units.) As the sequence is played, the time/frame indicator moves along the graph. The Timing Graph can also be used in a real time fashion in which the graph is created on the fly.

Area Zoom

In some instances, there is tremendous value in being able to ascertain more detail on the image. If the user places an area on the screen, it is possible to provide a zoom display of the particular region of interest highlighted. In this example a small rectangle is placed in the center of the device under test. To the right there is a magnification of the pixels in the Region of Interest.

If we were to minimize the size of the ROI, we actually can see the temperature values of each pixel. Pretty powerful!

Using Regions of interest with Auto Gain Control

Altair allows the user to manually adjust the thermal brightness and contrast should they choose. Alternately there are a series of AGC (Auto Gain Control) algorithms that can be applied. These range from user definable, to max-min, to full scale to histogram. When the user selects one of these AGC algorithms, Altair will automatically scale the scene.

However, it is common that the target. of interest in the image may not occupy the entire image. In addition it is possible that some other event occurs in the field of view that is much hotter or colder than the actual target of interest. If an ROI is defined over the target of interest, the AGC algorithm can be applied only inside of the ROI. Therefore any other temperature event happening outside the ROI will be ignored and the AGC will apply only within the ROI.

In this image, AGC is applied to the entire image. The muzzle flash of the gun is factored into the AGC algorithm therefore the temperature span is stretched wider. There is little to no contrast on the shooter.

However, if we did want to actually capture data on the shooter, we can simply touch the ROI and that activates the AGC within the ROI only. In this case the muzzle flash is ignored.

Transferring ROI Data to 3rd Party Programs

Any of the data captured from the ROI can be copied and pasted into third party programs, such as Excel, Word, Powerpoint, MatLab, etc. Whether it is the historgram, point, Timing Graph, or Area Zoom, the full data is captured. In addition, the "header" information (including time, frame, sequence duration, etc.) is also displayed. Using the area from the gun image above, here is an example of what the data would look like in Excel. Each pixel's coordinate and value is diplayed.

As you can see, with only one simple measurement tool, Altair can provide significant amounts of data. This only scratches the surface of the capabilities of Altair.