UV infrared image fusion combines two kinds of information that are useful when inspecting energized power equipment: ultraviolet emissions associated with corona discharge and infrared imagery that shows thermal conditions. By registering the images into a shared view, an inspection system can help operators relate a suspected discharge location to the surrounding equipment and its temperature. This approach supports online condition monitoring without relying solely on inspections performed after equipment is shut down.
Why combine ultraviolet and infrared inspection?
Outdoor electrical equipment operates under demanding conditions, including high temperature, high electrical stress, and sustained loading. Deterioration can lead to faults or accidents, so identifying developing problems before failure is an important maintenance objective. Traditional preventive inspection may require an outage. Yet equipment in a stopped state may not behave exactly as it does in service, which can affect what an inspection reveals.
Ultraviolet and infrared imaging provide different evidence. A partial discharge can emit ultraviolet radiation, and detecting emissions in the solar-blind ultraviolet band can help identify corona activity. Infrared imaging, in turn, provides information about the thermal state of equipment. Used together, the modalities can show where ultraviolet activity is observed while also displaying temperature information for the corresponding area.
Neither image should be treated as a complete diagnosis on its own. The combined view is an inspection aid: it brings the observed discharge signal and thermal context together so that personnel can assess the equipment and decide whether further investigation is warranted.
How solar-blind UV imaging detects corona
Ultraviolet radiation spans a broad spectral range, but atmospheric absorption makes wavelengths below approximately 280 nm largely absent at the ground. This region is known as the solar-blind band. A detector designed for this band can therefore observe relevant ultraviolet emissions against a background with very little solar ultraviolet contribution.
The described system uses a solar-blind UV camera with a detection band of 240–280 nm. Its imaging module uses an intensified charge-coupled device (ICCD). In simplified terms, incoming ultraviolet radiation reaches a photosensitive cathode, where it produces photoelectrons. A microchannel plate amplifies the electrons, which then strike a phosphor screen and create photons. An optical fiber interface carries that image to a visible-light CCD camera for capture.
Adjusting the ICCD gain changes the detector’s sensitivity within its operating range. Gain therefore needs to be set appropriately for the inspection, balancing the ability to detect ultraviolet activity with the quality of the resulting image. In the reported sensitivity calibration, a monochromator supplied ultraviolet light from 240 to 280 nm in 10 nm steps. The system’s observed photon-number trend followed that of a comparison UV detector.
Why image registration matters
A UV image and an infrared image cannot simply be overlaid without accounting for their different fields of view and pixel scales. In the described setup, the UV camera has a narrower field of view than the infrared camera. As a result, a single UV pixel and a single infrared pixel do not represent the same physical area, and the UV image covers a smaller portion of the scene.
Image registration aligns corresponding features so that the fused display places ultraviolet activity in the correct position relative to the infrared scene. The system establishes this relationship using geometric calibration and image features. A calibration plate measuring 1,000 by 1,000 mm, with regularly spaced 5 mm holes, provides known points for comparison. The plate is imaged in both bands, with the holes illuminated using a fluorescent lamp.
Because the calibration plate’s circular features appear elliptical in the UV image, the system identifies their edges and estimates each ellipse’s center using a least-squares method. Corresponding centers in the UV and infrared images are then used to calculate the translation between them. The corner locations help determine image scaling. The calibration also checks alignment in horizontal, vertical, and diagonal directions; lens distortion may be treated as negligible where it is sufficiently small.
This process establishes the geometric relationship needed to place the smaller UV image over the appropriate region of the infrared image. Registration quality is central to interpretation: a fused display is useful only when the two bands’ corresponding locations are aligned well enough for the operator to relate them.
From camera data to an inspection display
The system sends image data from both cameras to a host computer. The UV image is displayed in grayscale, while the infrared image is processed using a predefined temperature scale and displayed in pseudo-color. The software can show the individual images as well as the fused view, and images can be stored for later reference. It also identifies the highest temperature point in the infrared image and plots that temperature information.
Keeping the cameras’ optical paths fixed supports the registration process. In operation, the host software receives the two image streams, applies the registration and fusion steps, and displays the target area. The operator can use the separate views to inspect each modality and the combined view to understand their spatial relationship.
Practical implications for power equipment monitoring
For utilities and teams responsible for energized assets, dual-band imaging offers a way to bring discharge-location information and thermal context into one inspection workflow. Its value depends on more than the cameras themselves: spectral sensitivity, gain adjustment, geometric calibration, stable camera alignment, and clear presentation all contribute to whether the images can be interpreted usefully.
The reported system and field measurements indicate that the approach can support detection of power equipment faults and provide an online inspection option. In practice, fused images should be considered alongside established inspection procedures and the expertise of maintenance personnel, rather than as a substitute for engineering judgment.
By combining solar-blind UV detection with infrared temperature imaging, UV infrared image fusion helps make two complementary observations available in a single spatial context. For condition-monitoring programs, the practical takeaway is to pay close attention to calibration and registration as well as detection: accurate alignment is what turns separate spectral images into a coherent view for assessing corona activity and equipment condition.