Measuring Motion with Imaging Software
- Created: Saturday, 01 March 2014
High-speed photography is as much of an engineering tool as an oscilloscope, spectrum analyzer, or logic analyzer. The photographic technique enables us to visualize and analyze motion, especially motion that is too fast for the human eye or conventional cameras to perceive.
For decades, the defense, scientific and research, and industrial communities have used high-speed digital video files to extract motion and measure moving objects. As high-speed digital cameras continue to make advancements in recording speeds, sensitivity, and resolutions so must the motion analysis software used to extract the data they record, such as:
• Data that allows the defense community to examine the speed, angle, and angular speed of a shock wave from an explosive device.
• Information automotive engineers require to evaluate the safety and effectiveness of an airbag design (by determining the time, speed, and angle as the airbag full deploys).
• Analysis of human locomotion by measuring the angle of a bending knee joint, or the compression that the knee joint endures while running.
• Manufacturer needs, such as the angular speed (revolutions per minute) that a new hard drive motor can spin without causing damage to the disk, or the effect that a golf club’s angle of impact will have on the rotational speed of a golf ball (see Figure 1).
The ability to analyze all of this data quickly and accurately inherently decreases product development time, and more importantly, reduces research and development expenditures. Of course, extracting the information from a digital high-speed video is only as good as the tools used to accomplish it.
2-D motion analysis tools, such as Vision Research PCC (Phantom Camera Control) software, calculate the valuable data. With today’s software, the end-user can perform timing, position, distance, velocity, angle, and angular speed measurements, and track multiple points or objects to compute and graph their XYcoordinates, speed, or acceleration. PCC, for example, provides several edge detection algorithms and image processing tools to improve the measurement process. The motion analysis system harmonizes measured data with images. In this article, we will review the various measurement capabilities.
Units of Measurement
Units of Measurement specify the computing and reporting unit for distance, speed, acceleration, angle, and angular speed measurements. Establishing a measurement scale is required to set a specified number of pixels in the image equal to a scale unit size, such as millimeters, meters, inches, feet, or pixels. To define a measurement scale, the analyst selects two points on the image with a known scale, then specifies that scale size. Once created, all measurements are computed and displayed using the scale unit. If no measurement scale exists, the default scale will be 1 pixel = 1 pixel.
To perform timing measurements accurately, a time stamp (date and time) is embedded into every captured frame. The PCC software function, for example, calculates the time difference between two event frames (start / end of an event) or from the captured image being displayed to the trigger (t0) frame automatically.
Coordinate measurements are calculated from an Origin point pixel, by default the top-left corner of the image; however, the Origin can be changed when performing measurements. Each coordinate consists of two numbers (x1, y1), indicating the position of a pixel in the image on the two-dimensional plane from the Origin point.
Distance, Angle, Speed
Instant measurement tools analyze launch speed, angle, and angular speed, or the revolutions of a rotating object. The software capabilities allow engineers developing large-caliber weaponry, for example, to analyze the effect that the design of the shell has on the projectile trajectory; launch speed and angle determine the optimal performance. Manufacturers of scientific equipment, like an anemometer, can also determine the best size and type of motor to enhance their product; angular speed (rotational measurements) can be performed on the motors used to generate an electric current as they rotate (see Figure 2).
PCC measures the distance from the Origin point to a selected point, and the angle made by the Origin and Ox axis of the selected point, using the Distance and Angle and Speed: Origin + 1 Point instant measurement tool.
Given the coordinates of two points on the image plane, the distance (d) between the points is calculated using the following formula:
If the Origin and the selected point are on the same image, PCC will calculate distance and angle only; however, if the Origin and the selected point are on different frames, the software also calculates speed and angular speed.
Speed is calculated using the formula: s= d/dt , where d = measured distance, and dt = [time of the point frame] – [time of the origin frame] if point and origin are on different frames. Angular speed is calculated using the formula: as=a/dt , where a = measured angle, and dt = [time of the point frame] – [time of the origin frame] if point and origin are on different frames.
The Angle and Angular Speed: 3 Points instant measurement tool from PCC calculates the angle made by three points (two lines with a common reference point), while the Angle and Angular Speed: 4 Points tool calculates the angle formed by four points (two lines without a common reference point).
If all the points are in the same image when performing three- or four-point measurements, the software only calculates the angle. In order to compute angular speed, the first point and the reference point must be on the same image, while the second point (in a three-point measurement) must be on a different image.
The angle speed measurement tools are exceptional when analyzing a rotating object, such as a projectile rotating in mid-air. The tools can also be applied to other sciences; for example, manufacturers who develop unbreakable glass windows may want to study the way that a 2 × 4 tumbles in hurricane-force winds. Engineers may also use the methods to develop stabilizers for aircraft.
To investigate the effect environmental conditions may have on the recorded data, a National Instruments™ USB- or M- Series Data Acquisition (DAQ) module can also be used to acquire data from a wide range of sensors, and synchronize the data with slow-motion video images recorded on a Phantom camera, using Phantom Camera Control software (see Figure 3).
Phantom camera control and Cine playback, analysis, and measurements can be customized to meet specific test protocols using the Phantom System Developer Kit (SDK) for LabVIEW (Laboratory Virtual Instrument Engineering Workbench) or MATLAB (matrix laboratory) drivers.
The LabVIEW SDK contains visual instrument (VI) files needed to call Phantom SDK functions from LabVIEW, various utilities, and demo applications. The SDK uses the LabVIEW interface to shared libraries to call functions from Phantom libraries. The MATLAB SDK contains header files needed to call Phantom SDK functions from MATLAB, function wrappers, a simple object-oriented layer, and demo scripts.
SDKs allow, for example, automotive manufacturers to create command line scripts to control a Phantom camera directly from a computer, or run it in a Graphical User Interface specifically designed to perform or analyze airbag tests without having to use PCC.
Digital high-speed video has been and continues to be a useful test and measurement tool.
Imaging software performs complicated mathematical calculations to compute distance, speed, angle, angular speed, or acceleration measurements of single or multiple points from 2D images.
The ability to perform these calculations with clicks of a mouse button allows engineers, researchers, and developers to significantly reduce research and development time, thereby increasing productivity. The software enables more precise and accurate analysis of ballistics, explosions, weapon development, trajectory, biomechanics, sport performance, flow analysis, crash, combustion, and stress studies.
This article was written by Frank J. Mazella, Learning Products Manager at Vision Research (Wayne, NJ). For more information, visit http://info.hotims.com/49743-141.