The speckle reduction on the multidimensional television signals is an important task in measurement machine vision systems. The greatest error due to speckle is detected in triangulation optical sensors if measuring objects have a periodic structure. When measuring such objects, a random interference pattern determines speckle noise. Speckle noise can be reduced by a two-camera profile sensor. Cameras capture the multidimensional image of an object in different angles and the interference pattern on the images will also be different. This allows removing it from the image. The first feature of processing is signal superposition in conditions of distortion. This problem is solved by preprocessing. The second problem is speed processing of superposition. It is solved using pyramid transformation and optical flow estimation. The developed speckle noise reducing technique was tested on multidimensional television signals with images of drill-pipe threads. In the comparison of single-camera profile sensors the error of the shape estimation of the object decreased from 0.20 mm to 0.05 mm.
We describe iteration algorithm for multidimensional television signals superposition of optical triangulation sensor for speckle reducing, which distorts the measurements of object counter. The object of research is the drill-pipe thread. The measurements of such objects are very sensitive to the speckle. The estimation error can be about 0.20 mm and higher. But value of tolerance is about 0.05 мм. The developed method for speckle reducing includes the preliminary stage of superposition of two multidimensional signals. The location of speckle on signals is different. The different location is defined by object offset relatively the laser of optical triangulation sensor. The multidimensional television signals superposition allows estimating the location of the speckle noise on images. Removing of the image region with the speckle noise reduces the error of object contour estimation. In issue of television signals processing the one of important problem has been distortion of multidimensional signals. It connects with distortion of triangulation sensor optical elements. The method of superposition on background of optical distortions and speckle noise is described in this article.
We describe the iteration algorithm for multidimensional signals superposition of optical triangulation sensor. The multidimensional signals link by transformation. This transformation includes offsets, scale, and rotation. The processing of multidimensional signals has characterized some features, which make superposition difficult. This is an irregular sampling step and no matching between points of processing signals. We developed the algorithm on-base iteration procedure for this issue. The procedure includes the offsets estimation in the Cartesian system, and it includes the scale and rotates in the log-polar coordinate system. The numerical simulation allows the estimate of superposition parameters error. The simulation shows that superposition parameters error of developed algorithm likes error by brute force algorithm, but the developed algorithm is faster. The developed algorithm can use in a system that works in real-time.
The estimation of the offset, angle rotation, and other deformation of multidimensional optical video signals is an important task at intelligence video systems. One of the main tasks for estimation of the multidimensional signal deformation is increasing the speed of data processing. The television signals characterized a significant volume of data. This property is a constraint for applying brute force methods. Those methods are the universal solution for estimating unknown parameters of the parametric model. This research describes the issue of measurement offset, rotation, and scale with additive and multiplicative noise for spatial-temporal superposition of television images. The reduction of time processing is provided by the iteration procedure of unknown parameters estimation. This procedure consists of approximation by separate estimate offsets, scale, and rotates.
The article describes an important task for restore object shape, which is measured by triangulation optical sensor. The restored object shape (object contour) makes by spatio-temporal multidimensional signal processing. The data of object contour is received at close moments in time. The core of spatio-temporal multidimensional signal processing bases on the hypothesis that object shape changes slightly at close moments in time. The right restore object shape is provided by the superposition of contour. The superposition algorithm is based on the iteration procedure. The processing contours have matched closer to each other with each new iteration. The first contour is the etalon multidimensional signal. The second contour is a measured signal at close moments in time. Superposition makes by feature points. The feature point of the etalon signal is the contour point, the feature point of the measured signal is the intersection of a normal line (it is calculated by etalon signal) and measured contour. The computer simulation and verification test show improvement result measurement by the developed algorithm.
Multidimensional optical signals from triangulation sensors are described the highest level of information. In particular, it allows defining the object counter. The counter is allowed measuring solid deforming. The original superposition method was shown at this article. It includes the defining reference points at a counter of object etalon and calculates the line equations as approximation points of measured counter in area of reference points. Superposition by reference points and line equations allow solving the problem preliminary decomposition measured counter to fragments.
The method of recognition rail fishplate was shown at this article. The method was developed for track geometry car. The automation the process of measuring is the one important thing, which improves the railway safety. The check of recognition object takes much less time than the viewing of all track video data. So the information about railway condition is quickly given to railway expert. The article is interested for developer of video data processing.
This article was described the task offset and rotate measurement for video signal superposition. The developed algorithm is generalization of Lucas-Kanade model. Another name of this model is optical flow algorithm. The original algorithm allows to rate offsets parameters, but developed algorithm also allows to rate rotate angle. The concept of superposition has based on Taylor series of signal and sinus and cosines replacement on approximate function.
The triangulation optical sensors are widespread at industry. In particular, they are used at development of railway diagnostic system. The output signal of this sensor is multidimensional point sequences. At every single moment the signal defines the counter of target object. In the case it is rail counter. The one of important problem is method‘s development for recognition railway objects is the railway switch. The recognition method was shown at article. It is based on measurement side point of rail. At the area of switch triangulation optical sensors recognition at the same time two rails. The processing of railway parts allows recognition this situation.
This article has been described some features of processing multidimensional optical signals for measurement solid deforming. The object shape is measured by optical triangulation scanner. The solid deforming is defined by compare measured profile and underwear profile of aim object. The developing algorithm for profiles comparison has been shown in this research. It consists by two stages. The first stage is splitting set of measuring point’s profile on subset. The subset is certain part of object, such as straight line and circular arc. The second stage is superposition part of objects and corresponding line’s equation. The sample of superposition measured rail profile and underwear rail profile, compassion with existing method of solid deforming has been shown in conclusion.
KEYWORDS: Scanners, Multidimensional signal processing, Distortion, Signal processing, Video surveillance, Photosensitizer targeting, Distance measurement
This article was described some features of processing multidimensional optical video surveillance signals for triangulation scanner linearization algorithm. So, exactly: the obtainment source data, measurement point’s coordinate by physical location of target object, measurement coordinate of their points at photosensitive matrix coordinate system, the rating the operator transform from photosensitive matrix coordinate system to physical coordinate system. The result of linearization quality by the root mean square value was showed at conclusion. RMS value was calculated by real and measuring point’s coordinate of some part of scanner measuring filed.
KEYWORDS: Imaging systems, Multidimensional signal processing, Video surveillance, Signal processing, Optical signal processing, Diagnostics, Cameras, Algorithm development, Video, Telecommunications
This article has been described some features of processing multidimensional optical signals, which have been taken in real condition video surveillance of the railway carriages. Exactly: there are the choice region of image for speed measuring, the restore missed measurements, compensation of the camera vibration. The result is speed measurement algorithm of the extensional objects (the object does not contain in video frame fully), which is used for simplify operator working: control carriages braking on hump yard, making consignor list, control conditions of carriages detention.
KEYWORDS: Video surveillance, Optical signal processing, Optical components, Diagnostics, Detection and tracking algorithms, Signal processing, Video, Video processing, Image processing, Image filtering, Algorithm development
Processing of optical signals, which are received from CCD sensors of video cameras, allows to extend the functionality
of video surveillance systems. Traditional video surveillance systems are used for saving, transmitting and preprocessing
of the video content from the controlled objects. Video signal processing by analytics systems allows to get
more information about object’s location and movement, the flow of technological processes and to measure other
parameters. For example, the signal processing of video surveillance systems, installed on carriage-laboratories, are used
for getting information about certain parameters of the railways. Two algorithms for video processing, allowing
recognition of pedestrian crossings of the railways, as well as location measurement of the so-called “Anchor Marks”
used to control the mechanical stresses of continuous welded rail track are described in this article. The algorithms are
based on the principle of determining the region of interest (ROI), and then the analysis of the fragments inside this ROI.
Access to the requested content is limited to institutions that have purchased or subscribe to SPIE eBooks.
You are receiving this notice because your organization may not have SPIE eBooks access.*
*Shibboleth/Open Athens users─please
sign in
to access your institution's subscriptions.
To obtain this item, you may purchase the complete book in print or electronic format on
SPIE.org.
INSTITUTIONAL Select your institution to access the SPIE Digital Library.
PERSONAL Sign in with your SPIE account to access your personal subscriptions or to use specific features such as save to my library, sign up for alerts, save searches, etc.