Radiometric Vs RGB Thermal Imagery

In the world of remote sensing and thermal imaging, two terms often come up: radiometric and RGB (Red Green Blue) thermal imagery. While both are used to capture and analyze thermal data, one seems to have proven a distinct advantage.

RGB Thermal Imagery

RGB thermal imagery combines traditional RGB (red, green, blue) color information with thermal data to create a single image that overlays thermal information onto a visible light image. This technique is often used in thermal cameras that feature a built-in RGB camera, allowing users to see both thermal and visual information simultaneously.  Most RGB thermal imagery is captured in grey scale which only provides a maximum of 255 shades of grey available for interpretation.

In the photo below, the hot spot to the left is a BBQ with a temperature of over 100 Celsius where as the hotspot on the right is the pilot and visual observer which are around zero degrees Celsius relative to ambient temperature . Looking at both spots at a glance they appear to each be white hot. An AI may be able to extrapolate more from the different shades but is ultimately limited without actual temperature data.

However, an advantage of RGB thermal imagery is its ease of interpretation and small file size.  By overlaying thermal data onto a visible light image, users can quickly identify thermal anomalies and correlate them with specific objects or areas in the scene. This makes RGB thermal imagery a good for applications such as home inspection, solar load monitoring and industrial monitoring where the color shade difference visible to the human eye is sufficient to assess the data.

Radiometric Thermal Imagery

Radiometric thermal imagery refers to images captured by thermal cameras that record temperature data for each pixel in the image. This means that each pixel represents a specific temperature value, allowing for precise temperature measurements and analysis.  Most consumer level radiometric thermal IR cameras will have a resolution of 640x512 pixels.  This means by obtaining data for each pixel we now have 327,680 individual data points that can be analyzed.

Another key advantage of radiometric thermal imagery is its ability quantify temperature differences with high accuracy. This makes it ideal for applications such as search and rescue and wildfire hotspoting where a small significant heat signature could be missed by the human eye but captured by even just a few pixels. When the same image as before is processed using its radiometric data we see now that we only identify the left hotspot (BBQ) as a positive because the temperature threshold prevents the colder pilot and VO from being tagged. Temperature thresholds can be increased further to isolate even more strong positives.

The graph below illustrates temperature spikes within the radiometric image.

Software used for processing: USR FireTrak

Because of the amount of data captured by radiometric thermal imagery it currently requires post-processing to extrapolate the data. This doesn’t always allow cloud processing but greatly increases the flexibility of data analysis after the fact such as the ability to adjust temperature ranges, apply different color palettes, and extract temperature data for further analysis, providing valuable insights into thermal patterns and anomalies.

Key Differences

• Data Interpretation: Radiometric thermal imagery provides precise temperature data for each pixel, allowing for detailed analysis. RGB thermal imagery, on the other hand, overlays thermal data onto a visible light image, providing a more intuitive way to interpret thermal information.
  

• Post-Processing: Radiometric thermal imagery offers more flexibility in post-processing, as analysts can adjust temperature ranges and apply different color palettes. RGB thermal imagery is more limited in post-processing options, as the thermal data is overlaid onto a visual image color spectrum.
  

• Applications: Radiometric thermal imagery is ideal for applications that require precise temperature measurements and detailed analysis allowing.  RGB thermal imagery is more suited for applications that require quick interpretation of thermal data.

In conclusion, both radiometric and RGB thermal imagery have their advantages and are suited for different applications though when It comes wildfire or search and rescue operations radiometric imagery has taken the lead on thermal data analysis.

Click Here for more information on the FireTrak Software shown here.