Motion Scan and Data Collection Methods for Electro-Optic System Testing | Aerotech Whitepaper


Aerotech | Whitepaper

Motion Scan



Motion Scan and Data Collection Methods for Electro-Optic System Testing


By: Tom Markel

Aerospace Strategic Accounts Manager, Aerotech


This article discusses different scanning patterns and approaches that can be used to measure an electro-optic (EO) sensor’s performance. Measurements of sensor resolution, distortion, field-of-view (FOV) size, minimum resolvable contrast and minimum resolvable temperature (in the case of IR cameras) are all examples of performance metrics that are measured with these tests.


Executing scan tests with an automated motion system allows for capturing many data points that characterize sensor performance for development and production testing. Automating these tests with a very accurate system allows pass/fail testing while establishing production sensor in-spec baselines and performance trends that can be used to find quality and manufacturing issues. These automated tests can even improve sensor uniformity and accuracy by using the measured image metrology data to calibrate out optical errors by digitally correcting the EO image data in real time.


Single Point Target Scanning


One of the simplest and most effective tests of an EO imaging system is to uniformly move a high-contrast target through the sensor FOV and collect the line-of-sight position vs sensor image data at each location. This process starts with programming point-to-point steps over a defined pattern – typically a square, rectangle, circle or crosshair – through the sensor FOV. At each test point, data are captured from the EO sensor simultaneously with the line of sight motion encoder positions.


The step pattern could be square or rectangular raster scans, corner scans or circular scans of uniform rings centered in the FOV.


Figure 1 shows an example of a unidirectional raster scan pattern. This scan pattern is often used to perform EO sensor calibration when the highest accuracy is necessary. Using a unidirectional scanning approach removes motion system hysteresis and measurement backlash. One disadvantage of this scanning approach is longer calibration and measurement time.

Figure 1. Unidirectional, single-point target pattern step sequence.


Figure 2 shows an example of a bidirectional raster scan pattern. This scan pattern is often used when throughput is critical, as testing time is minimized. Also, this pattern is a good test approach if the sensor under test has a lower resolution than the motion system hysteresis and backlash. One potential pitfall of this approach is that motion system hysteresis and backlash can lower overall accuracy.

Figure 2. Bidirectional, single-point target pattern step sequence.


Figure 3 shows an example of a raster scan pattern that includes corner measurements. This scan pattern is good for quickly locating the image center, checking the image skewness and showing distortions and symmetry of sensor optical alignment with the imager, which helps to identify assembly errors. For applications where the EO sensor is mounted on a pan/tilt gimbal, this scan pattern can be used to measure the orthogonality of the EO sensor to the gimbal motion axes. Correction for that alignment error is either performed in the mechanical setup or in software.

Figure 3. Raster scan pattern with image checks at the corners.


Figure 4 shows an example of a circular scan using concentric circles and measurement points at45° increments. Circular scan paths are frequently used to measure the optical telescope’s lens alignment during initial assembly. This test is often repeated in a temperature chamber to determine alignment sensitivity to the temperature ranges that the sensor is exposed to in operation. Finally, this scan pattern can be used for laser spot power distribution measurements and can help identify local hot spots or power voids caused by design or manufacturing issues. A tight spiral scan pattern can also achieve measurement results similar to those of concentric circle scans.

Figure 4. Circular scan pattern with concentric circles and measurement points located at 45° increments.


All of the aforementioned scan patterns can be used to measure lens distortion error. At the beginning of each scan pattern, the motion system aligns the target to the center of the FOV. The XY pixel location and motion encoder positions are stored as a center reference. After the entire scan is completed, the motion encoder positions are compared to the unit-under-test (UUT)sensed target XY pixel locations to determine the lens distortion error. A sensor correction lookup table can be generated from this data. Verification of this error map is then performed by rerunning the motion system through the programmed scan pattern with the error map correction file loaded in software.




Resolution Target or Array Target Scanning


Instead of a single point target scan, a two-dimensional resolution target or array of target fiducial scan be used to cover a larger area of the FOV at each motion step. This approach increases image target measurement locations for a single motion step by instantaneously recording a 2D array of target XY pixel locations in the EO sensor. It also decreases the overall number of motion steps across the FOV. Figure 5 shows an example of this two-dimensional area raster scan approach.

Figure 5. Two-dimensional area raster scan illustration.


An advantage of this approach is that measurement time is reduced. However, post-processing is slightly more complex since it requires stitching images and combining data.


The same scan approaches that were previously illustrated in Figures 1 through 4 can be used with this overlapping area approach. This technique is good for calibrating very large FOV sensors such as satellite imagers. It also is a good scan method to survey a large area with a gimbal-mounted, high-resolution sensor that has a narrow FOV. By adding image processing to the system, overlapping scanning can also be used to locate new or changing objects in the gimbal’s field-of-regard.


Dynamic Sensor Testing


Sensor dynamics and image processing performance can be tested by moving objects at constant or changing rates through a field of view. The test object is typically a small vertical bar or thin plate which would pass between a uniform background and the sensor. This is sometimes referred to as a slit test. Various motion devices such as rotary and linear stages can be used depending on the scan frequency, target size and scan lengths. The sensor imaging bandwidth and instantaneous imaging resolution can be tested by increasing motion speed until the slit is no longer visible.


The same motion scan profiles shown previously can be used, but instead of stopping at each measurement location, the data are captured on the fly using tight synchronization between the UUT and motion system. Measurement accuracy may be slightly lower due to the dynamics of both the motion system and sensor. However, the main advantages are faster testing time and higher throughput.




Data Collection for Dynamic Testing


Ensuring that the measurement data is exactly matched to the measurement position can be challenging in high-resolution dynamic testing. Faster scan rates are advantageous for higher throughput, but tight position synchronization of the UUT data with the motion data is necessary.


Some advanced motion controllers have built-in, low-latency triggering that can be used to achieve this tight synchronization. Aerotech motion controllers include a position-based, low-latency output trigger signal called Position Synchronized Output (PSO). PSO allows the user to specify a vector distance in three-dimensional space for triggering data collection. The output can be triggered off of a commanded distance or actual position feedback, further enhancing the synchronization fidelity.


Using built-in controller features like PSO eliminates the need for purchasing expensive high-speed hardware data acquisition systems or developing custom data acquisition software.




In this article, the advantages and disadvantages of various motion scan techniques are discussed. Using these scan techniques along with the test setups and actuation techniques discussed in related articles will allow an EO test engineer to make more informed decisions regarding an effective test strategy.





Hong Kong’s Largest Aircraft Maintenance Service Provider Achieves 300% Improvement in Integration Time with Boomi 
Blue laser sintering material for 3D printing from igus ensures even greater food safety