For automating deburring of cast parts, this paper proposes a general method for estimating burr height using 3D vision sensor that is robust to missing data in the scans and sensor noise. Specifically, we present a novel data-driven method that learns features that can be used to align clean CAD models from a workpiece database to the noisy and incomplete geometry of a RGBD scan. Using the learned features with Random sample consensus (RANSAC) for CAD to scan registration, learned features improve registration result as compared to traditional approaches by (translation error ( $$\Delta $$ 18.47 mm) and rotation error( $$\Delta 43 ^\circ $$ )) and accuracy(35%) respectively. Furthermore, a 3D-vision based automatic burr detection and height estimation technique is presented. The estimated burr heights were verified and compared with measurements from a high resolution industrial CT scanning machine. Together with registration, our burr height estimation approach is able to estimate burr height similar to high resolution CT scans with Z-statistic value ( $$z=0.279$$ ).

We consider the case of robotic bin picking of reflective steel parts, using a structured light 3D camera as a depth imaging device. In this paper, we present a new method for bin picking, based on a dual-resolution convolutional neural network trained entirely in a simulated environment. The dual-resolution network consists of a high resolution focus network to compute the grasp and a low resolution context network to avoid local collisions. The reflectivity of the steel parts result in depth images that have a lot of missing data. To take this into account, training of the neural net is done by domain randomization on a large set of synthetic depth images that simulate the missing data problems of the real depth images. We demonstrate both in simulation and in a real-world test that our method can perform bin picking of reflective steel parts.

This article deals with Short Wave Infra-Red (SWIR) and Long Wave Infra-Red (LWIR) imaging sensors for detecting icebergs in harsh metocean conditions. Field data acquired during the Statoil Offshore Newfoundland Research Expedition 2015 (ONRE15) is analyzed. The analysis is supported by a numerical modelling study which aims at simulating the optical properties of ice and water combined with the radiation transfer in the Infra Red.

Concrete Innovation Centre (COIN), is currently working with the development of a specification tool for concrete surfaces cast against smooth formwork. The tool covers greyscale and distribution of pores. Greyscale and pores shall be classified separately, which implies two separate systems for these parameters. Both classification systems will probably include four classes with demands. All classes will be illustrated with pictorial examples. Both systems shall include a project specific class for which demands can be made for a specific project. The systems shall also include a class without any demands.

This paper presents a NIR multispectral imaging system in the region 760–1040nm that employs non-contact optics to allow interactance measurements to be made. By using interactance, absorption of the light is increased and measurement of inhomogeneous or intact products is possible. Applications from the fish and meat industry are presented here

The objective of the current study was to analyze and develop some important automation steps in a superchilling process line. In order to control the product quality there is a need for online measurements of ice fraction and distribution in inhomogeneous products. Moreover, automatic handling of such superchilled products is currently commercially unavailable. The current study presents a new method for monitoring and handling superchilled product of varying form and consistency. Observation of the shift in the water absorption peak, measured by near-infrared spectroscopy (NIR) transflection mode, was used to determine the ice level in superchilled salmon, scanning approximately 1.5cm into the fillets. The salmon fillets were stored on ice at 0°C for 5–7days before superchilling at −24°C to target ice contents of 10%, 15% and 30%. Online NIR measurements of ice fraction showed promising results, with a low prediction error of 2.5%. The storage study confirmed former quality results with a microbiological shelf life of 15–17days with only minor differences in values for drip loss and water-holding capacity between superchilled and chilled samples.

A novel system for on-line measurement of fat content in inhomogeneous pork trimmings is presented. The system allows near infrared (NIR) energy to interact with the meat using non-contact optics while it is travelling in large plastic boxes on a conveyor belt. A comparison was made between the log of the inverse of the interactance NIR spectra [log(1/ T)], standard normal variate (SNV) and extended multiplicative signal correction (EMSC) as techniques for the correction of physical light scattering due to colour and textural differences, height variation and temperature fluctuations, depending on whether the meat was warm-cut or cold-cut. EMSC gave the best prediction results; a robust partial least squares regression using two factors resulted in a root mean square error ( RMSEP) of 1.9% on 20 kg batches of inhomogeneous meat trimmings. The model was fully tested twice in an on-line environment at a slaughter house and performed with a RMSEP of 3.4% for a fat range of 8–55% in the first industrial trial and 2.82% in the second industrial trial.

The physical performance characteristics of a prototype scanning digital mammography (DM) system have been investigated. The I-ImaS system utilises CMOS MAPS technology promoting on-chip data processing; consequently statistical analysis is therefore achievable in real-time for the purpose of exposure modulation via a feedback mechanism during the image acquisition procedure. The imager employs a dual array of twenty CMOS APS sensing devices each individually coupled to a 100 mum thick thallium doped structured CsI scintillator. The X-ray performance of the sensors was characterised where the presampled modulation transfer function (MTF), normalised noise power spectrum (NNPS), and the detective quantum efficiency (DQE) was determined. The presampled MTF was measured utilising the slit technique and was found to be 0.1 at 6 lp/mm. The NNPS measured utilising a W/Al target/filter combination hardened with 38 mm PMMA was seen to decrease with increasing exposure as expected and the manifesting DQE was 0.30 at close to zero spatial frequency at an exposure of 1.75 mR. Preliminary image stitching of the individual steps acquired from the scanning system is presented. A conventionally acquired image that is without the implementation of beam modulation or off-line intelligence is compared and contrasted to an intelligently off-line processed image. Results indicate the implementation of real-time intelligence into the image acquisition phase of digital mammography is foreseeable.

I-ImaS (Intelligent Imaging Sensors) is a European project aiming to produce real-time adaptive X-ray imaging systems using Monolithic Active Pixel Sensors (MAPS) to create images with maximum diagnostic information within given dose constraints. Initial systems concentrate on mammography and cephalography.In our system, the exposure in each image region is optimised and the beam intensity is a function of tissue thickness and attenuation, and also of local physical and statistical parameters in the image. Using a linear array of detectors, the system will perform on-line analysis of the image during the scan, followed by optimisation of the X-ray intensity to obtain the maximum diagnostic information from the region of interest while minimising exposure of diagnostically less important regions.This paper presents preliminary images obtained with a small area CMOS detector developed for this application. Wedge systems were used to modulate the beam intensity during breast and dental imaging using suitable X-ray spectra. The sensitive imaging area of the sensor is 512×32 pixels 32×32μm2 in size. The sensors' X-ray sensitivity was increased by coupling to a structured CsI(Tl) scintillator. In order to develop the I-ImaS prototype, the on-line data analysis and data acquisition control are based on custom-developed electronics using multiple FPGAs.Images of both breast tissues and jaw samples were acquired and different exposure optimisation algorithms applied. Results are very promising since the average dose has been reduced to around 60% of the dose delivered by conventional imaging systems without decrease in the visibility of details.

Objective: A prediction model was developed to simulate an underexposed radiograph to that which is optimally exposed. The objective of the study was to evaluate if radiographic landmarks were equally observed in lateral cephalometric radiographs processed with and without the prediction model.Methods: Using a digital imaging system (Orthopantomograph (R) OC-100D), test radiographs of a human skull phantom covered with simulated soft tissue were obtained using different exposure settings. The optimal radiograph was subsequently established. The under-exposed raw data radiographs were then processed in two sets. In one set, the radiographs were first simulated from the optimally exposed raw data radiograph using the prediction model and then processed with the default settings of the proprietary software that was used to control the imaging system used. In the second set, the radiographs were processed only with the default settings of the proprietary software. Two monitors were employed to simultaneously display the radiographs. Six observers subjectively compared the fourteen most frequently used landmarks on both radiographs.Results: Compared with the non-predicted radiographs, in the predicted radiographs subjective impression of the radiographic landmarks was more frequently considered the same or better than the optimal radiographs. The difference was statistically significant.Conclusion: The subjective impression of the landmarks in the predicted radiographs is better than that in the non-predicted radiographs.

I-ImaS (Intelligent Imaging Sensors) is a European project aiming to produce new, intelligent x-my imaging systems using novel APS sensors to create optimal diagnostic images. Initial systems have been constructed for medical imaging; specifically mammography and dental encephalography. However, the I-ImaS system concept could be applied to all areas of x-ray imaging, including homeland security and industrial QA.The I-ImaS system intelligence is implemented by the use of APS technology and FPGAs, allowing real-time analysis of data during image acquisition. This gives the system the capability to perform as an on-the-fly adaptive imaging system, with the potential to create images with maximum diagnostic information within given dose constraints.The I-ImaS system uses a scanning linear array of scintillator-coupled 1.5-D CMOS Active Pixel Sensors to create a full 2-D x-ray image of an object. This paper describes the parameters considered when choosing the scintillator elements of the detectors. A study of the positioning of the sensors to form a linear detector is also considered, along with a discussion of the potential losses in image quality associated with creating a linear sensor by tiling many smaller sensors.Preliminary results show that the detectors have sufficient performance to be used successfully in the initial mammographic and encephalographic I-ImaS systems that are currently under construction.

96 Volume 11, Number 2, 2008 The basic principle for dental radiographic examination is ALARA (As Low As Reasonably Achievable) to the patient dose1. This indicates that patient exposure should be kept as low as possible while producing a radiograph of sufficient quality for diagnosis. This is especially important in young people and adolescents, as risks from x-ray radiation may be higher for these patients2. Although radiation exposure has been reduced through the use of rare-earth intensifying screens3, different collimator shapes and digital imaging systems4-7, this could be improved by adaptively exposing tissues or organs that are of interest to the radiologist. Recently, a project named I-ImaS (Intelligent Imaging Sensors) was funded by the European Commission under 1 Department of Oral and Maxillofacial Radiology, ACTA, Amsterdam, The Netherlands. 2 Department of Medical Physics and Biomedical Engineering, University College London, London, UK. 3 Manager of ANCO S.A. development team, Athens, Greece. 4 Rutherford Appleton Laboratory, CCLRC, Oxford, UK. 5 Department of Informatics and Telecommunications, University of Athens, Athens, Greece. 6 Division of Electronics and Cybernetics, SINTEF, Oslo, Norway. 7 Physics Department, University of Ioannina, Ioannina, Greece. 8 Medical Physics Group, University of Trieste, Trieste, Italy.

I-ImaS (Intelligent Imaging Sensors) is a European project which has designed and developed a new adaptive X-ray imaging system using on-line exposure control, to create locally optimized images. The I-ImaS system allows for real-time image analysis during acquisition, thus enabling real-time exposure adjustment. This adaptive imaging system has the potential of creating images with optimal information within a given dose constraint and to acquire optimally exposed images of objects with variable density during one scan. In this paper we present the control system and results from initial tests on mammographic and encephalographic images. Furthermore, algorithms for visualization of the resulting images, consisting of unevenly exposed image regions, are developed and tested. The preliminary results show that the same image quality can be achieved at 30-70% lower dose using the I-ImaS system compared to conventional mammography systems.

I-ImaS (Intelligent Imaging Sensors) is a European project aiming to produce adaptive x-ray imaging systems using Monolithic Active Pixel Sensors (MAPS) to create optimal diagnostic images. Initial systems concentrate on mammography and cephalography. The on-chip intelligence available to MAPS technology will allow real-time analysis of data during image acquisition, giving the capability to build a truly adaptive imaging system with the potential to create images with maximum diagnostic information within given dose constraints. In our system, the exposure in each image region is optimized and the beam intensity is a function not only of tissue thickness and attenuation, but also of local physical and statistical parameters found in the image itself. Using a linear array of detectors with on-chip intelligence, the system will perform an on-line analysis of the image during the scan and then will optimize the X-ray intensity in order to obtain the maximum diagnostic information from the region of interest while minimizing exposure of less important, or simply less dense, regions. This paper summarizes the testing of the sensors and their electronics carried out using synchrotron radiation, x-ray sources and optical measurements. The sensors are tiled to form a 1.5D linear array. These have been characterised and appropriate correction techniques formulated to take into account misalignments between individual sensors. Full testing of the mammography and cephalography I-ImaS prototypes is now underway and the system intelligence is constantly being upgraded through iterative testing in order to obtain the optimal algorithms and settings.

Re-invented in the early 1990s, on both sides of the Atlantic, Monolithic Active Pixel Sensors (MAPS) in a CMOS technology are today the most sold solid-state imaging devices, overtaking the traditional technology of Charge-Coupled Devices (CCD). The slow uptake of CMOS MAPS started with low-end applications, for example web-cams, and is slowly pervading the high-end applications, for example in prosumer digital cameras. Higher specifications are required for scientific applications: very low noise, high speed, high dynamic range, large format and radiation hardness are some of these requirements.This paper will present a brief overview of the CMOS Image Sensor technology and of the requirements for scientific applications. As an example, a sensor for X-ray imaging will be presented. This sensor was developed within a European FP6 Consortium, intelligent imaging sensors (I-ImaS). (c) 2007 Elsevier B.V. All rights reserved.

We present results from a study where we segment fish in images captured within fish cages. The ultimate goal is to use this information to extract the weight distribution of the fish within the cages. Statistical shape knowledge is added to a Mumford-Shah functional defining the image energy. The fish shape is represented explicitly by a polygonal curve, and the energy minimization is done by gradient descent. The images represent many challenges with a highly cluttered background, inhomogeneous lighting and several overlapping objects. We obtain good segmentation results for silhouette-like images containing relatively few fish. In this case, the fish appear dark on a light background and the image energy is well behaved. In cases with more difficult lighting conditions the contours evolve slowly and often get trapped in local minima

I-ImaS is a European project aiming to produce new, intelligent x-ray imaging systems using novel APS sensors to create optimal diagnostic images. Initial systems concentrate on mammography and encephalography. Later development will yield systems for other types of radiography such as industrial QA and homeland security.The I-ImaS system intelligence, due to APS technology and FPGAs, allows real-time analysis of data during image acquisition, giving the capability to build a truly adaptive imaging system with the potential to create images with maximum diagnostic information within given dose constraints.A companion paper deals with the DAQ system and preliminary characterization. This paper considers the laboratory x-ray characterization of the detector elements of the I-ImaS system. The characterization of the sensors when tiled to form a strip detector will be discussed, along with the appropriate correction techniques formulated to take into account the misalignments between individual sensors within the array.Preliminary results show that the detectors have sufficient performance to be used successfully in the initial mammographic and encephalographic I-ImaS systems under construction and this paper will further discuss the testing of these systems and the iterative processes used for intelligence upgrade in order to obtain the optimal algorithms and settings.

We have developed a 1.5D CMOS active pixel sensor to be used in conjunction with a scintillator for X-ray imaging. Within the Intelligent Imaging Sensors (I-ImaS) project, multiple sensors will be aligned to form a line-scanning system and its performance evaluated with respect to existing sensors in other digital radiography systems. Each sensor contains a 512×32 array of pixels and the electronics to convert the collected amount of charge to a digital output value. These include programmable gain amplifiers (PGAs) and analogue-to-digital converters (ADCs). The gain of the PGA can be switched between one or two, to increase the sensitivity for smaller collected charge; the ADC is a 14-bit successive approximation with a sampling rate of 1.25MHz. The ASIC includes a programmable column fixed pattern noise mitigation circuit and a digitally controllable pixel reset mode block. Here we will describe the sensor design and the expected performance.