High speed cameras are used to capture pictures of very rapid events that would be impossible to capture with normal cameras. As such, they are extensively used also for aerospace testing purposes. Industry experts assess the use of high speed cameras in the domain of aerospace testing, the specific requirements of the aerospace industry with regard to high speed cameras, the factors to consider when selecting a high speed camera and artificial intelligence as an area of integration.
A typical flight test instrumentation high speed camera is capable of taking images at speeds of 200 frames per second or higher and these are collected using a global shutter, which means that the entire image is captured at the same moment in time, affirms Ben Kupferschmidt, senior product line manager for network and ground solutions at Curtiss-Wright Defense Solutions. “The result is a clear image without motion blur, despite the very short exposure time. High speed camera imagery can then be played back at slow speeds to allow engineers to perform data analysis on complex events involving rapid motion,” he says. “In the aerospace industry, high speed cameras are useful for a number of flight test instrumentation (FTI) applications. High speed cameras can also help with confirming proper operation for any complex moving surfaces on the aircraft, for example aerial refuelling and de-icing systems. By using multiple high speed cameras, it is possible to capture multiple angles of the same event. The imagery data from these different angles can be combined to help generate a 3D model of the target event.”
High speed cameras are capable of capturing super-fast events by recording thousands, or even millions of frames per second, affirm Anders Kalldahl and James Engel of Phantom Ametek. “Most events are in the millisecond or microsecond range in duration; indeed, we sometimes refer to high speed cameras as ‘time microscopes’. A traditional microscope enlarges objects too small to be seen by the naked eye, while a high speed camera expands the time of an event that is too fast for the naked eye and the brain to process,” they say. “In the aerospace industry, high speed cameras are commonly used in high speed wind tunnels for aerodynamics and shockwave studies, component stress and performance testing using techniques such as digital image correlation (DIC), flow visualization through particle image velocimetry (PIV), in-flight stores release testing, and rocket motor testing.”
Rob Huculak and Daniel Castillo of the National Institute for Aviation Research (NIAR) at Wichita State University affirm that high speed cameras provide valuable information from a wide range of experimental tests in the aerospace industry. “They capture, at large frame rates, fast-paced events, allowing test engineers to have clear exploitable visuals. This provides insight at the microsecond scale on how deformations take place, how velocities change, or failures occur during dynamic events,” they affirm. “A major benefit comes from advanced photogrammetry software which enables the direct measurement of displacements of test objects in a high speed test; this software requires a calibration to be done for each camera, lens, and focus setting. It corrects for parallax, perspective, depth, and scale factor in order to output the correct displacement, for example that of an occupant kinematics in seat testing. One of the most advanced areas where high speed cameras are used is DIC where a pair of cameras is calibrated and the test images are analysed simultaneously to determine the 3D coordinates and displacements of the object of interest.”
In the aerospace industry, high speed cameras are used for the research and development of engines (spray and combustion), materials (to improve strength, to develop new materials, to lighten weight), welding (optimizing the welding method), and visualizing high speed mechanical structures, affirms Alex Kiendl of MESSRING, a distributor for nac cameras for the European and Indian market. “Cameras are used not only for visualization but also for analysis by using various image processing technologies, such as two-color method/two-color ratio method for temperature analysis, DIC (for deformation measurement), PIV (for flow measurement), the Shlieren method (for visualizing air density change), and motion analysis,” he says.
High speed cameras can also be coupled to thermal imaging. Thermal imaging cameras convert infrared radiation into a visual image that depicts temperature variations across an object or scene; this allows the cameras to make non-contact measurements of an object’s temperature for data acquisition, analysis and reporting, according to Jerry Beeney, director of global business development at Teledyne FLIR. “The process of using an infrared camera for data viewing, recording, analysis and reporting is often referred to as ‘thermography’,” he affirms. “High speed thermal cameras often refer to cameras which can achieve frame rates in excess of 100 frames per second, with some being able to reach maximum frame rates of greater than 50,000 frames per second. This allows the cameras to accurately measure temperatures of fast thermal transients and moving targets.”
High speed thermal cameras are routinely used in the research, development and testing phases within all sectors of the aerospace industry to verify and improve the thermal performance of critical aviation electronics, points out Beeney. “When mated with the appropriate microscope lenses, high speed thermal cameras can detect small thermal anomalies in individual components of integrated circuits. They are also used in material testing to detect potential defects in critical components in airframes and propulsion systems,” he says. “These non-destructive testing (NDT) processes usually involve active thermography techniques where an excitation source creates a thermal difference which the thermal imaging camera can detect and evaluate for potential weaknesses or deficiencies.”
The range of applications within the aerospace industry is quite broad, and so are the specific requirements of the aviation industry about high speed cameras. “Taking NIAR AVET as an example, we possess several high speed cameras with a maximum frame rate of 2,000 frames per second and few more cameras capable of going as high as 480,000 frames per second,” say Huculak and Castillo. “Due to the nature of the aerospace industry, there is constant push towards maximizing the quality and amount of information gathered for any given test. Hence, manufacturers are continuously looking to improve the capabilities of their equipment catalog.”
According to Kupferschmidt, the aviation industry needs high speed cameras that are rugged and capable of surviving the extreme environments in which aerospace vehicles operate. These requirements include the need to function across large temperature ranges and high G-forces, and the cameras also need to be relatively small, low power, and capable of operating autonomously. “Ideally, the cameras should be triggered automatically when the target event happens. This ensures that the timing is correct, since a key limitation of high speed cameras is the fact that they have a limited amount of internal storage for images,” he says.
Multiple cameras are typically mounted at different locations on board the aircraft, providing different fields of view of the target event, observes Kupferschmidt. “In order to combine the images from multiple cameras into a single model for data analysis, it is vital that the time on each camera is in sync. This enables the engineers to be confident that the images from different cameras are aligned in time,” he affirms. “Requirements for high speed camera applications are sent to high speed camera manufacturers using the normal request for proposal process. Most customers are looking for a commercial off-the-shelf (COTS) solution that has the flexibility to accommodate their unique needs.”
Kalldahl and Engel affirm that the specific requirements of the aviation industry include frame rates from 1000 frames per second to 75,000 frames per second at a minimum of 1-megapixel resolution, and in the case of DIC testing, 4-megapixel sensors are required for best results. Furthermore, many tests are performed with less-than-optimal lighting conditions so sensors with high sensitivity to the available light are needed. “This high sensitivity is especially important since low exposure times (fast shutter speeds) are usually required to eliminate motion blur,” they say. “These requirements are usually communicated through a meeting with the test engineers and an on-site visit to evaluate the test parameters and environmental conditions in order to match their specific requirements to a particular camera model. Years of experience with aerospace testing has led to sensor development that addresses the specific needs of aerospace test engineers.”
High speed thermal cameras are continuing to improve in both pixel resolution and maximum frame rates, according to Beeney. “The aviation industry in particular is interested in higher pixel resolutions at higher frame rates since it can help them to get more measurement pixels on the article being tested at high frame rates,” he affirms. “The increase in pixel density on the measurement target improves the accuracy of the non-contact temperature measurement while also providing a more detailed understanding of the thermal gradients on objects of interest. The aviation industry also requires high speed thermal cameras to utilize industry standard connections and protocols for data and camera control.”
Decision Making Factors
There are many factors that go into selecting a high speed camera. According to Kupferschmidt, the most critical initial decision is to determine the required frame rate and resolution for the images. He specifies that most aerospace flight test instrumentation applications need between 200 and 500 frames per second and that the customers typically want a camera resolution of 1280×1024. “Other key factors include the size, weight and power requirements for the specific test. In particular, high speed cameras often have to fit in tight spaces on board the test aircraft,” he says. “Another key factor to consider is the triggering method for the high speed camera. Ideally, aerospace customers seek flexibility in their cameras that support multiple options for triggering. The overall system architecture and system integration is vital to a successful high speed camera test; customers want a fully integrated system where all of the different pieces work together seamlessly.”
Among the factors that should be assessed when deciding the type of high speed camera to buy, there must be always a balance between required frame rate versus resolution, minimum shutter speed/available light versus sensor sensitivity and record time, according to Kalldahl and Engel. “Image quality parameters, such as noise and dynamic range, need to be considered because the images are often used for measurements and data analysis,” they affirm. “To a lesser extent, form factor and ruggedization sometimes play a role when space and/or weight is limited or if the camera is exposed to harsh conditions such as an exterior wing mount.”
The purchase decision starts with the application; having a well-defined and constrained application facilitates the equipment selection, observe Huculak and Castillo. “In general, the primary factors for deciding on the type of high speed camera to buy are different. One of this is the frame rate; that refers to the number of images that a camera captures over a period of time. The optimal frame rate setting for a test should be determined from the velocity and the overall magnitude of the displacement of the object of interest: it should be enough to capture the event in thorough detail,” they say. “Another factor is the resolution, i.e., the number of pixels that constitute the captured image. The higher the resolution the more details or features can be distinguished on the image. For high speed cameras, frame rate and resolution are inversely proportional. That is, a high frame rate setting will lead to a lower maximum resolution setting and vice versa. It is up to the test engineers to determine the optimal balance between the two.”
A key point is also the cost. In general, the cost of a high speed camera is primarily driven by the image sensor combination of frame rate, resolution and light sensitivity capabilities, affirm Huculak and Castillo. “A camera with a sensor capable of recording full HD images at high frame rates will result in a significant investment. Other factors to be evaluated are internal memory capability, external memory options, available trigger input and delay options, synchronization, overall equipment dimensions, weight, g-rating, etc.”
When it comes to selecting thermal imaging camera, the first point is about what temperatures one is expected to measure, observes Beeney. “The second key item is how fast one needs to capture the data. The third factor to consider is the size and distance to the target of interest. The fourth point refers to the kinds of temperature analysis and report generation that are needed. The fifth and last item has to do with what additional accessories are required to complete the final solution,” he says. “Concerning the latter item, project equipment requirements may extend beyond the need for an infrared camera and software. For example, one may need a protective enclosure to use the camera in a harsh or demanding environment. Safety requirements may force one to work miles away from the camera, necessitating a remote operation system.”
Today there are high speed streaming cameras available, observe Kalldahl and Engel. “Traditional cameras capture images in a finite amount on internal RAM, and then download and analyse them after the event has occurred, while streaming cameras send raw image data directly to a frame buffer in a PC. Like a traditional camera those images can be stored on a DVR for longer recording sessions, but more often users are sending those images through an FPGA, CPU, and GPU,” they say. “This allows them to do real-time high speed analysis, decision-making, and adjustments at thousands of frames per second. For example, adjustments to an airframe or other component mid-flight to compensate for an anomaly or simply to change the characteristic of the test.”
According to Kupferschmidt, artificial intelligence (AI) is applicable to high speed camera data analysis. “One application is path tracking within the imagery data. For high speed camera tests that involve fast moving projectiles, it is useful to be able to visualize the result in 3D and to track the path of the moving objects,” he says. “Comparing multiple tests that occurred at different times is a promising AI application. When viewing imagery from multiple tests, it can be hard for engineers to notice subtle differences. AI software, however, should be able to identify these subtle differences and potentially discover problems with the test aircraft’ systems.”
Indeed, the use of AI and high speed cameras is an area of open research and development at the time, and major photogrammetry software developers have started to successfully implement AI, according to Huculak and Castillo. “The results concern primarily image recognition and autonomous tracking of unique targets and features resulting in shorter image analysis time and effort for engineers and researchers,” they say.
According to Kiendl, AI could be implemented more often together with high speed video images in the future. “One potential market is the quality judgement at production line and one of our trail in this market is a real-time welding quality judgement,” he affirms.
According to Beeney, AI is one type of computer vision currently being with imaging systems, including high speed thermal imaging cameras, to automatically extract important data from images and videos. “These algorithms can save time and improve repeatability by quickly finding relevant information and is particularly useful in thermal imaging where imagery can be less intuitive, he concludes.