TY - JOUR
AU - Schmalzl,, Jörg
AB - Abstract To position arbitrary outdoor measurements, the most common methods are the use of global positioning system (GPS) technology or distance tracking from reference locations. The system to be introduced here consists of a combination of an optical distance tracking system and a civil, autonomous GPS receiver. Through the application of a recursive filter, both components support and correct each other mutually. The system is intended as an autonomous, real-time and low-cost alternative to expensive differential GPS solutions in geophysical field surveying. The combination approach features a very good relative and decent absolute positioning precision. First measurements using the two-coil geo-electromagnetics instrument Geonics EM31 in combination with the developed positioning system show promising results. environmental science, Earth science, instrumentation, measurement Introduction For the economy of ground-based, small-scale geophysical field surveying, rapid, high-quality data acquisition is desirable. Therefore, not only have adequate instruments to be utilized, but also a precise and fast data positioning method is essential. Position information can be obtained by several methods. The most common are distance tracking from reference locations or the use of GPS (global positioning system) technology. Tracking wheel odometers are precise, yet positioning is relative, as the quantity being measured is a distance, not as absolute point in space. Furthermore, measurement errors accumulate with covered distance and, due to slippage or loss of contact, they cannot be used on any ground. GPS position tracking using autonomous, civil GPS receivers is absolute, but it is, for the application in geophysical surveying, generally not accurate enough. To improve GPS positioning, expensive (both in purchase and work effort) differential GPS (DGPS) solutions have to be applied. The positioning system to be introduced here consists of a combination of a common GPS receiver and a two-dimensional odometer based on an optical mouse sensor. Both systems mutually support each other, as GPS position integrity is improved by the additional displacement information and accumulating errors in distance tracking are compensated by the incorporation of absolute GPS positions. Optical flow tracking Besides their obvious deployment in optical computer mice, optical mouse sensors are extensively used as two-dimensional odometers in the field of mobile robotics (Palacin et al2005, Baek et al2005, Dille et al2009). To use this technique also in geophysical surveying applications, a prototype of an outdoor optical tracking system is introduced here. An optical mouse sensor in principle consists of a low-resolution camera with a high frame rate and an image processing unit. Here, the Agilent ADNS3080 (Agilent Technologies GmbH 2005) high-performance optical mouse sensor (figure 1) is used. By comparing consecutive frames of a sequence, the image processor can determine the two-dimensional projection of, in general, three-dimensional relative displacements between the camera and the observed surface. To do this, sensor algorithms compute the optical flow on the camera image plane. Output information is a displacement (dx, dy) in units of pixels. Figure 1 Open in new tabDownload slide The Agilent ADNS3080 optical mouse sensor (left) features a frame rate of 6400 fps. A sensor frame (right) has a resolution of 30 × 30 pixels (Agilent Technologies GmbH 2005). The depicted image is made with the sensor being mounted on a CCTV lens with an elevation of 60 cm above a computer keyboard. Figure 1 Open in new tabDownload slide The Agilent ADNS3080 optical mouse sensor (left) features a frame rate of 6400 fps. A sensor frame (right) has a resolution of 30 × 30 pixels (Agilent Technologies GmbH 2005). The depicted image is made with the sensor being mounted on a CCTV lens with an elevation of 60 cm above a computer keyboard. Optical flow is a two-dimensional vector field in the image plane of the camera. A vector proportional to the displacement between consecutive frames of a pixel's intensity value is computed for every pixel on the image plane (figure 2). Using optical flow algorithms (Horn and Schunck 1980, Lucas and Kanade 1980, Beauchemin and Barron 1995), rotations and translations of images or subimages can thus be determined. For the application in optical mice, only a global translation of the whole image is of interest. Figure 2 Open in new tabDownload slide Schematic of the optical flow (c) calculated from two consecutive frames (a) and (b) of a sequence. Indicated are a translation (top) and a rotation (bottom) of subimages. Figure 2 Open in new tabDownload slide Schematic of the optical flow (c) calculated from two consecutive frames (a) and (b) of a sequence. Indicated are a translation (top) and a rotation (bottom) of subimages. Due to the assumed operation on even surfaces close to the sensor lens, the standard optical mouse configuration is not practical in outdoor applications. Therefore, the optical sensor is mounted on a 16 mm CCTV (close circuit television) camera lens to to able to increase its elevation and thus its footprint size. To credibly detect displacements, the observed surface has to be sufficiently textured. Otherwise, algorithms cannot keep track of recognizable features in consecutive frames. As a measure of intensity variance within an image, the ADNS3080 reports a surface quality value. Experience shows that the sensor, in combination with the camera lens, outdoors and in daylight detects sufficiently high surface quality values on any ground to reliably determine displacements. To transform the sensor's displacement in units of pixels into a real-world displacement in units of metres, information on the distance from the lens to the observed surface has to be available because the footprint size of a camera normal to a surface and thus the ‘real size’ of a pixel on its image plane depends linearly on that distance (figures 3(a) and (b)). Therefore, the ultrasonic ranger Devantech SRF021 is installed next to the optical sensor. It determines the distance to a reflector by measuring the sonic pulse echo runtime. A height-dependent conversion factor s, which maps pixel displacements to real-world distances, is then assessed experimentally. Figure 3 Open in new tabDownload slide Dependence of the optical mouse sensor's displacement measurements on its alignment relative to the surface. Moving the sensor in alignments (a), (b) and (c) at a given distance parallel to the surface would result in three different displacement measurements. Additional distance to the surface and orientation measurements thus are necessary to correct the measured values. Figure 3 Open in new tabDownload slide Dependence of the optical mouse sensor's displacement measurements on its alignment relative to the surface. Moving the sensor in alignments (a), (b) and (c) at a given distance parallel to the surface would result in three different displacement measurements. Additional distance to the surface and orientation measurements thus are necessary to correct the measured values. Sensors are installed strictly horizontal on a mobile device and for different sensor heights h, a known straight distance is covered. The known distance is divided by the optical sensor's pixel displacement and plotted against the height. The desired conversion function is the interpolation to the measurements (figure 4). Figure 4 Open in new tabDownload slide Dependence of the conversion factor s on height h, which transfers the optical sensor's displacement from units of pixels to units of metres. Figure 4 Open in new tabDownload slide Dependence of the conversion factor s on height h, which transfers the optical sensor's displacement from units of pixels to units of metres. Displacements can now be expressed in units of metres, yet only in a local, sensor-defined coordinate system. Furthermore, the sensor is unable to detect any rotations. To overcome this problem, the three-dimensional magnetic field and acceleration sensor Honeywell HMC6343 (Honeywell International Inc., 2008) is installed. This sensor is able to detect a tilt compensated compass bearing and pitch and roll angles. With this information, on the one hand, errors resulting from image distortions due to sensor dipping (figure 3(c)) can be computed and compensated. On the other hand, the sensor's displacement (dx, dy) in its local coordinate system can be transferred to a geographic displacement (de, dn) in a Cartesian geographic map projection (figure 5): Figure 5 Open in new tabDownload slide Mapping the optical sensors's local coordinate system to a geographic, Cartesian projection. Figure 5 Open in new tabDownload slide Mapping the optical sensors's local coordinate system to a geographic, Cartesian projection. With the applied sensors, it is now possible to determine a rotation-independent displacement in real-world units on a geographic map projection's coordinate system (the magnetic north direction in a first approximation is assumed to be the same as the geographic north direction). The installation of the sensors is depicted in figure 6. Figure 6 Open in new tabDownload slide The installation of the sensors. (a) Bottom: an ultrasonic ranger and a camera lens. (b) Top: display for the state of health, displacement, distance to surface and orientation information. (c) Inside view with sensors and a microcontroller. The microcontroller transfers sensor data to an USB port. Figure 6 Open in new tabDownload slide The installation of the sensors. (a) Bottom: an ultrasonic ranger and a camera lens. (b) Top: display for the state of health, displacement, distance to surface and orientation information. (c) Inside view with sensors and a microcontroller. The microcontroller transfers sensor data to an USB port. The precision of displacement measurements depends on the installation of the sensors. With firmly horizontal adjustment in constant height and orientation on a mobile device and on a straight profile (compass sensor thus not needed), errors are well beyond 1%. When manually and freely carried around, variations in the sensor's height and orientation are inevitable. Errors detected by first test measurements in that configuration are in the order of about 6%. GPS tracking Position determination with civil and autonomous GPS receivers inherits biases of several metres. Figure 7 shows the details of a test measurement carried out with a modern SiRFStarIII GPS receiver. For the whole timespan of that measurement, the receiver remained stationary. Figure 7 Open in new tabDownload slide Test measurements with a stationary SiRFstarIII GPS receiver. The timespan covered is 16 h with a record interval of 15 s. The mean value has been subtracted from all records. The standard deviation is 2.02 m and the maximum deviation is 10.04 m. Figure 7 Open in new tabDownload slide Test measurements with a stationary SiRFstarIII GPS receiver. The timespan covered is 16 h with a record interval of 15 s. The mean value has been subtracted from all records. The standard deviation is 2.02 m and the maximum deviation is 10.04 m. As a stand-alone positioning system for use in geophysical field surveying, these receivers, in general, are too inaccurate. The major drawback here is the error-prone relative positioning of consecutive GPS measurements (see also figure 9). An interpolation calculation on associated geophysical field measurements will thus weight adjacent data points inaccurately. This leads to blurred out structures in the obtained maps (figure 10). To improve positioning precision, additional spatial information provided by secondary sources is inevitable. This information may originate from other GPS receivers (DGPS), or from any system gathering relative or absolute position data. DGPS methods use two or more spatially proximate GPS receivers. Significant errors in position determination, for example the ionospheric signal runtime delays or satellite ephemeris and clock errors, affect these receivers in a similar manner. Stationary base receivers on known positions are thus able to calculate correction data with which positions acquired by rover receivers are adjusted (El-Rabbany 2002). Correction data are incorporated either in real time or in post-processing. Real-time differential solutions rely on expensive communication hardware and constant signal availability. Post-processing is time consuming and data visualizations are not possible until single records are spatialized. As an alternative solution, here, optical distance tracking is to be introduced as a secondary spatial information source to improve GPS positioning. Combination of GPS and optical flow sensors GPS measurements of autonomous, civil receivers are absolute, yet insufficiently precise. Measurement errors of the optical distance tracking system are relatively low for small displacements, but grow considerably with covered distance. When combined, on the one hand, GPS position integrity can be improved by the additional displacement information. On the other hand, relative errors in distance tracking can be compensated by the incorporation of absolute GPS position information. To combine measurements from the GPS and an optical sensor, both have to share a single-coordinate system. With the aid of the electronic compass, optical sensor displacement is available in a geographic, Cartesian coordinate system. Most common GPS receivers report position information in the NMEA-0183 format. Given that output, a location is specified by geographic latitude and longitude in degrees. To map the geographic latitude and longitude to a geographic and Cartesian coordinate system, such as for example the UTM (Universal Transverse Mercator) system, there are commonly defined projection algorithms available (Kienast et al1994). Displacements and GPS positions are combined here using a recursive filtering process. The filtering approach to be introduced here shares the predictor–corrector and data-weighting ideas of the statistically optimal Kalman filtering (Maybeck 1979, Welch and Bishop 2006, Zarchan 2005). Based on a starting position the optical displacement is used to predict a new position and a weighted GPS measurement corrects that position. The weight is determined by the relative measurement errors of displacement ΔD and GPS ΔG. The GPS error is assumed to be constant and twice the standard deviation (95% of measurements lie within that radius) of the test measurement introduced in figure 7: The displacement precision is assumed to be 6%. Its absolute value depends on the covered distance between two GPS measurements: Given that, the weight σ can be determined as and the recursive filtering equation becomes Application example For first performance tests, the newly developed positioning system is used in combination with the Geonics EM31. The EM31 is a two-coil geo-electromagnetics instrument. The optical tracking system is mounted on the instrument, together with a mobile palmtop computer with an internal GPS receiver (figure 8). Figure 8 Open in new tabDownload slide The optical positioning system and the palmtop computer mounted on the Geonics EM31. Figure 8 Open in new tabDownload slide The optical positioning system and the palmtop computer mounted on the Geonics EM31. The EM31 is mostly used for contact-free and fast detection of lateral ground conductivity variations near to the surface. The instrument can be operated by a single person. It is carried by the side at hip height by a shoulder strap (Geonics Ltd 1995). The palmtop is connected to the EM31 and the optical tracking system via USB, reading both the device's sensor data and its internal GPS data. During field measurements, the palmtop continuously computes the current position in real time based on the above filtering equations and logs it together with current EM31 measurements to its memory. As the optical system is mounted on the EM31 and carried at the operator's side, fluctuations in sensor alignment are inevitable. Results of a first test measurement of that combination approach are depicted in figure 9. Figure 9 Open in new tabDownload slide Test measurement with the combination approach of optical distance tracking and GPS positioning. North–south facing, parallel profiles with a spacing of 2 m and a length of 50 m are laid out. While covering these, sensors are carried at hip height mounted on the EM31. Raw GPS positions (green) and filtered positions (red) are logged with a frequency of 1 Hz. Figure 9 Open in new tabDownload slide Test measurement with the combination approach of optical distance tracking and GPS positioning. North–south facing, parallel profiles with a spacing of 2 m and a length of 50 m are laid out. While covering these, sensors are carried at hip height mounted on the EM31. Raw GPS positions (green) and filtered positions (red) are logged with a frequency of 1 Hz. Nevertheless, significant improvements are made, especially in relative positioning and even perpendicular to the profiles. With support of the optical tracking system, thus, considerably large fluctuations in the raw GPS measurements can be adjusted. The effects can be seen in interpolation maps of EM31 measurements originating at another test measurement on the military airbase in Fassberg, Germany. Figure 10 shows the same interpolation map section calculated for the same data set, where on the left data are positioned using the GPS alone and on the right supported by displacement information through the above filtering equation. Resolution of structures can be improved significantly and for many applications is thus sufficiently high. Figure 10 Open in new tabDownload slide Interpolation maps of EM31 measurements (calculated with ESRI ArcMap's spline interpolation). Both maps show the same map section and the same data, yet positioned by different methods. With data positioned by the GPS in combination with displacement information (right) a much higher resolution of structures is achieved than with data positioned by the GPS alone (left). Figure 10 Open in new tabDownload slide Interpolation maps of EM31 measurements (calculated with ESRI ArcMap's spline interpolation). Both maps show the same map section and the same data, yet positioned by different methods. With data positioned by the GPS in combination with displacement information (right) a much higher resolution of structures is achieved than with data positioned by the GPS alone (left). The integrity of the newly developed positioning method is verified by the successful return to the starting position (figure 9). Here only a small deviance is recognizable. Yet, the absolute positioning precision depends to a large extent on the precision of the starting value in the filtering process. In this test measurement, the starting value has been determined by a short-term average of 50 GPS measurements. Nevertheless, absolute precision is assumed to be only slightly better than that of a civil GPS receiver. For more convenient and more quantitative information on system precision and its dependence on, for example, surface, terrain, illumination, sensor fluctuations or GPS reception, further research will be done. Conclusion As a positioning system for geophysical field surveying, a combination of an optical distance tracking system and a civil, autonomous GPS receiver has been developed. Through the application of a recursive filtering approach, both components support and correct each other mutually. It is shown that the combination approach features a very good relative and a decent absolute positioning precision, even under considerable fluctuations in sensor alignment. The reached measurement resolution is, for many applications, thus sufficiently high. The system is intended as an autonomous and low-cost (the costs of the applied sensors, a microcontroller and a GPS receiver are about ⁠) alternative to expensive DGPS solutions in the field of mobile and small-scale geophysical surveying, especially in combination with the EM31. Without going into detail, there is a wide range of applications imaginable, in which DGPS corrections are not available and the provided stand-alone GPS positioning is insufficiently precise. Furthermore, if alignment fluctuations are reduced and the compass usage (and its biases) can be avoided, an application as ‘optical tracking wheel’ is self-suggesting. Possibly, autonomous tracking, with a precision in the order of conventional tracking wheel precisions, is achievable and the system can be used for measurement-triggering. The advantages here are evident, as optical tracking is contact-free and applicable even on grounds like, for example, fine sand or ice. Using focused artificial illumination, optical tracking can even be utilized in darkness. Considering the wide range of possible applications, further research will be done. Particularly, error analysis and targeted improvement of single components are the next steps. To be able to make more convenient statements on precision, further performance tests are planned. References Agilent Technologies GmbH . , 2005 ADNS-3080 High-performance Optical Mouse Sensor Data-Sheet Baek S , Park H , Lee S . , 2005 Mobile robot localization based on consecutive range sensor scanning and optical flow measurements Proc. 12th Conf. on Advanced Robotics (Department of Mechanical and System Design Engineering, Hongik University, Seoul) Beauchemin S S , Barron J L . , 1995 The computation of optical flow , ACM Computing Surveys (CSUR) London Department of Computer Science, University of Western Ontario Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Dille M , Grocholsky B , Singh S . , 2009 Outdoor downward-facing optical flow odometry with commodity sensors Proc. Field and Service Robotics (FSR '09) (Robotics Institute, Carnegie Mellon University, Pittsburgh) El-Rabbany A . , 2002 , Introduction to GPS: The Global Positioning System Boston, MA Artech House Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Geonics Ltd . , 1995 Geonics EM31-MK2 Operating Manual Honeywell International Inc. , 2008 HMC6343 3-Axis Compass with Algorithms Data-Sheet Horn B K P , Schunck B G . , 1980 , Determining Optical Flow (Artificial Intelligence Laboratory) Cambridge, MA Massachusetts Institute of Technology Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Kienast G , Hofmann-Wellenhof B , Lichtenegger H . , 1994 GPS in der Praxis, Abteilung für Landesvermessung und Landinformation, Technische Universität Graz (Berlin: Springer) Lucas B D , Kanade T . , 1980 An iterative image registration technique with an application to stereo vision Proc. 7th Int. Joint Conf. on Artificial Intelligence (Computer Science Department, Carnegie-Mellon University, Pittsburgh) Palacin J , Valgañon I , Pernia R . , 2005 The optical mouse for indoor mobile robot odometry measurement , Sensors Actuators , vol. 126 (pg. 141 - 7 )http://dx.doi.org/10.1016/j.sna.2005.09.01509244247 Google Scholar Crossref Search ADS WorldCat Maybeck P S . , 1979 , Stochastic Models, Estimation and Control vol 1 New York Academic chapter 1 Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Welch G , Bishop G . , 2006 , An Introduction to the Kalman Filter Chapel Hill, NC Department of Computer Science, University of North Carolina Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Zarchan P . , 2005 , Fundamentals of Kalman Filtering—A Practical Approach Lexington Lincoln Laboratory, Massachusetts Institute of Technology Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC Footnotes 1 " http://www.robot-electronics.co.uk/htm/srf02techI2C.htm © 2010 Nanjing Geophysical Research Institute
TI - Using an optical mouse sensor to track geophysical field measurements
JF - Journal of Geophysics and Engineering
DO - 10.1088/1742-2132/7/4/007
DA - 2010-12-01
UR - https://www.deepdyve.com/lp/oxford-university-press/using-an-optical-mouse-sensor-to-track-geophysical-field-measurements-u4pmYzR28P
SP - 404
VL - 7
IS - 4
DP - DeepDyve
ER -