No full-text available
To read the full-text of this research,
you can request a copy directly from the authors.
External shape measurement for industrial applications using artificial intelligence and optimised data fusion
Abstract
In this work, a priori information about the nominal geometry of a measured object and the operating principles of three optical form measurement techniques are used to improve system performance compared to using each technique individually. More specifically, we present a surface form measurement system which uses artificial intelligence and machine vision to enable the efficient combination of fringe projection, photogrammetry and deflectometry. The measurement system can identify the regions of an object that are optimally measured by each individual process and employs a measurement strategy in which the measurement systems are combined in concert to achieve a complete three-dimensional measurement of the object using a reference camera. The system has a target maximum permissible error of 50 μm and the prototype demonstrates the ability to measure complex geometries of additively manufactured objects, with a maximum size of (10 × 10 × 10) cm, with minimal user input.
... Additionally an accurate optical technique is being developed for form measurements of PAM 2 parts [19][20][21]. In the artificial intelligence (AI)enhanced data-fused optical measurement framework that is being developed [19] for quick post-process 3D shape measurements, a combi- the areal measurement. ...
... Additionally an accurate optical technique is being developed for form measurements of PAM 2 parts [19][20][21]. In the artificial intelligence (AI)enhanced data-fused optical measurement framework that is being developed [19] for quick post-process 3D shape measurements, a combi- the areal measurement. This type of setup reduces the cost and complexity of the system by combining all three measurement systems into one framework while enabling an expanded material application range. ...
PAM^2, which stands for Precision Additive Metal Manufacturing, is a European MSCA project in which 10 beneficiaries and 2 partners collaborate on improving the precision of metal Additive Manufacturing. Within this project, research is done for each process stage of AM, going from the design stage to modelling, fabricating, measuring and assessment. For each step we aim to progress the state of the art with a view on improving the final AM part precision and quality by implementing good precision engineering practice.
In this article we will list PAM^2’s detailed objectives, the envisioned approach and the results achieved after 1,5 years of research.
In the Light Controlled Factory part-to-part assembly and reduced weight will be enabled through the use of predictive fitting processes; low cost high accuracy reconfigurable tooling will be made possible by active compensation; improved control will allow accurate robotic machining; and quality will be improved through the use of traceable uncertainty based quality control throughout the production system. A number of challenges must be overcome before this vision will be realized; 1) controlling industrial robots for accurate machining; 2) compensation of measurements for thermal expansion; 3) Compensation of measurements for refractive index changes; 4) development of Embedded Metrology Tooling for in-tooling measurement and active tooling compensation; and 5) development of Software for the Planning and Control of Integrated Metrology Networks based on Quality Control with Uncertainty Evaluation and control systems for predictive processes. This paper describes how these challenges are being addressed, in particular the central challenge of developing large volume measurement process models within an integrated dimensional variation management (IDVM) system.
3D imaging sensors for the acquisition of three dimensional (3D) shapes have created, in recent years, a considerable degree of interest for a number of applications. The miniaturization and integration of the optical and electronic components used to build them have played a crucial role in the achievement of compactness, robustness and flexibility of the sensors. Today, several 3D sensors are available on the market, even in combination with other sensors in a “sensor fusion” approach. An importance equal to that of physical miniaturization has the portability of the measurements, via suitable interfaces, into software environments designed for their elaboration, e.g., CAD-CAM systems, virtual renders, and rapid prototyping tools. In this paper, following an overview of the state-of-art of 3D imaging sensors, a number of significant examples of their use are presented, with particular reference to industry, heritage, medicine, and criminal investigation applications.
Due to their speed, data density, and versatility, optical metrology tools play important roles in today's high-speed industrial manufacturing applications. Handbook of Optical Dimensional Metrology provides useful background information and practical examples to help readers understand and effectively use state-of-the-art optical metrology methods. The book first builds a foundation for evaluating optical measurement methods. It explores the many terms of optical metrology and compares it to other forms of metrology, such as mechanical gaging, highlighting the limitations and errors associated with each mode of measurement at a general level. This comparison is particularly helpful to current industry users who operate the most widely applied mechanical tools. The book then focuses on each application area of measurement, working down from large area to medium-sized to submicron measurements. It describes the measurement of large objects on the scale of buildings, the measurement of durable manufactured goods such as aircraft engines and appliances, and the measurement of fine features on the micron and nanometer scales. In each area, the book covers fast, coarse measures as well as the finest measurements possible. Best practices and practical examples for each technology aid readers in effectively using the methods. Requiring no prior expertise in optical dimensional metrology, this handbook helps engineers and quality specialists understand the capabilities and limitations of optical metrology methods. It also shows them how to successfully apply optical metrology to a vast array of current engineering and scientific problems.
We trained a large, deep convolutional neural network to classify the 1.2 million high-resolution images in the ImageNet LSVRC-2010 contest into the 1000 dif- ferent classes. On the test data, we achieved top-1 and top-5 error rates of 37.5% and 17.0% which is considerably better than the previous state-of-the-art. The neural network, which has 60 million parameters and 650,000 neurons, consists of five convolutional layers, some of which are followed by max-pooling layers, and three fully-connected layers with a final 1000-way softmax. To make training faster, we used non-saturating neurons and a very efficient GPU implemen- tation of the convolution operation. To reduce overfitting in the fully-connected layers we employed a recently-developed regularization method called dropout that proved to be very effective. We also entered a variant of this model in the ILSVRC-2012 competition and achieved a winning top-5 test error rate of 15.3%, compared to 26.2% achieved by the second-best entry
In this paper we show that, by using a photogrammetry system with and without laser speckle, a large range of additive manufacturing (AM) parts with different geometries, materials and post-processing textures can be measured to high accuracy. AM test artefacts have been produced in three materials: polymer powder bed fusion (nylon-12), metal powder bed fusion (Ti-6Al-4V) and polymer material extrusion (ABS plastic). Each test artefact was then measured with the photogrammetry system in both normal and laser speckle projection modes and the resulting point clouds compared with the artefact CAD model. The results show that laser speckle projection can result in a reduction of the point cloud standard deviation from the CAD data of up to 101 µm. A complex relationship with surface texture, artefact geometry and the laser speckle projection is also observed and discussed.
The scope of this review is to investigate the main post-process optical form measurement technologies available in industry today and to determine whether they are applicable to industrial-grade metal additive manufactured parts. An in-depth review of the operation of optical three-dimensional form measurement technologies applicable to metal additive manufacturing is presented, with a focus on their fundamental limitations. Looking into the future, some alternative candidate measurement technologies potentially applicable to metal additive manufacturing will be discussed, which either provide higher accuracy than currently-available techniques but lack measurement volume, or inversely, which operate in the appropriate measurement volume but are not currently accurate enough to be used for industrial measurement.
The main objective of this book is to present the basic theoretical principles and practical applications for the classical interferometric techniques and the most advanced methods in the field of modern fringe pattern analysis applied to optical metrology. A major novelty of this work is the presentation of a unified theoretical framework based on the Fourier description of phase shifting interferometry using the Frequency Transfer Function (FTF) along with the theory of Stochastic Process for the straightforward analysis and synthesis of phase shifting algorithms with desired properties such as spectral response, detuning and signal-to-noise robustness, harmonic rejection, etc.
A high-speed phase-shifting technique for 3D shape measurement has been developed and experimented. A Digital Micromirror Device (DMD) based video projector is used to project sinusoidal fringe patterns onto the object surface. Its red, green and blue (RGB) channels are programmed to be sinusoidal fringe patterns with 120-degree phase difference. The projector has a color filter with four segments, RGB and clear, which controls the color of the projected image. When a specific segment is in the light path, the DMD forms the corresponding image of that color. A 24-bit color image is formed when the RGB color channels are sequentially projected in a cycle time of 10 ms. When the color filter in the projector is removed, the DMD still projects the three color channels sequentially, but the actual projected images are now in grayscale. The timing control mechanism of the DMD is studied and a circuit is designed to provide an external timing signal for projection control. When a CCD camera used to capture the fringe patterns is synchronized with the DMD, the individual channels of RGB or the three phase-shifted fringe patterns can be captured in less than 10mn, thus making high- speed phase shifting possible. Experimental results demonstrated the feasibility of this technique for high-speed 3D shape measurement.
This paper describes a novel methodology for the reverse engineering of complex, free form surfaces, based on the integration
of the measurement information from a D vision sensor and a coordinate measuring machine (CMM). The aim is to reconstruct
the CAD model of objects of complex geometry with high accuracy and at the same time, rapidly exploiting the advantages deriving
from the use of both the optical and the mechanical sensors, with a minimum of human intervention. The combination is performed
at the level of measurement information, within a module for the intelligent aggregation of the information from optical and
mechanical sensors. Tools are developed for digitising, filtering, grouping and surface-fitting the 3D images of the vision
sensor and the point clouds from digitisation of the CMM. In the paper, the combined system is described, and two industrial
applications are presented.