Digital Image Analysis Techniques Research Articles

The development of an automated correction �procedure for digital photogrammetry for the study of wide, shallow, gravel-bed rivers

This paper develops an automated correction procedure for dealing with point errors associated with through-water photogrammetry, for application in the study of clear-water, shallow gravel-bed rivers. The procedure involves combining digital photogrammetry and image analysis techniques to: (i) correct for the effects of refraction at an air–water interface; and (ii) eliminate and reinterpolate points where the bed has not been ‘seen’. The correction procedure was applied to raw digital elevation models (DEMs) generated using digital photogrammetry from 1:3000 scale aerial photography of a small reach of the North Ashburton River, New Zealand. The accuracy of corrected and uncorrected DEMs is evaluated using an independent data set. A measure of ‘geomorphological usefulness’ as well as DEM external reliability is obtained from calculations of water depth distributions and mean bed level. Results show that digital photogrammetry, used in conjunction with image analysis techniques, can successfully be used for extracting high-resolution DEMs of gravel river beds. In exposed areas, errors are small and random, tending to cancel out over large numbers of points. Where water is shallow, and following correction, point elevation errors are statistically no different from those for exposed zones. In deeper water, despite an improvement following application of the correction procedure, elevation errors scale with water depth. The geomorphological potential of photogrammetric survey of large, gravel river beds is demonstrated by the ease and accuracy of calculations of water depth distribution (important for the assessment of a river's ecological and recreational characteristics) and mean bed level (important for the calculation of reach-scale sediment volumes). Copyright © 2000 John Wiley & Sons, Ltd.

Surveying for Migration Pathways in the Granitic Rock Using Nuclear Track Detectors, Autoradiography and Digital Image Analysis as an Aid To Construct the Basis for Heterogeneous Diffusion Modeling

ABSTRACTRadioelement migration within a rock matrix under natural long-term conditions is a complex process controlled by various parameters. Pure physical parameters such as porosity, hydraulic conductivity and diffusivity are usually sufficient to describe transport in well-defined laboratory systems. In natural rock matrices transport is influenced by physical pore properties such as pore size distribution, connectivity, tortuosity, constrictivity and petrological and chemical nature and charge on the fluid-rock interface. The overall characterization of heterogeneous rock structures is needed for the accurate heterogeneous diffusion modeling.Here we describe a method for the detection of μ-particles from uranium in cm-scale rock samples based on the analysis of the tracks formed in organic polymer, CR-39. On the other hand the uranium tracks were compared with the migration pathways and porosity distribution produced with the 14C-polymethylmethacrylate impregnation method (14C-PMMA). For analyzing mineral specific uranium occurrence and porosities the staining methods were used to produce the mineral map of the rock sample. Digital image analysis techniques were applied to the different cm-scale pictures of rock samples. Scanning electron microscopy and energy dispersive X-ray analysis (SEM/EDS) were performed in order both to study the pore apertures of grain boundaries and fissures in greater detail and to detect the uranium phases.The high uranium content was found to be congruent with the porous mineral phases; altered plagioclase and biotite grains, and the intra- and intergranular fissures detected with the 14C-PMMA technique. Plenty of microfractures transsecting potassium feldspar and quartz grains were filled with calcite together with precipitated uranium.

Rupture probability of coarse aggregate on fracture surface of concrete

In this work, the fracture surface of concrete was analyzed by the digital image analysis (DIA) technique, and the rupture probability of the coarse aggregate (RPCA) was used to represent the failure mode of the aggregate at the fracture surface. The relationships between the RPCA, the water-binder ratio (W/B), and the size and type of coarse aggregate were investigated. Preliminary results showed that: (1) RPCA increases with decreasing W/B of concrete. (2) RPCA of concrete with coarse aggregates having a maximum size of 16 mm is higher than that with other sizes of coarse aggregate. The influence of the size of coarse aggregate on the RPCA is more significant in high strength concrete than in normal strength concrete. (3) With reduction of W/B, the interfacial bond of concrete with crushed gravel improves more significantly than with round gravel according to their RPCA. (4) RPCA depends not only on the intrinsic strength, size, and shape but also on the reactivity of the coarse aggregate.

Influence of images recording height and crop growth stage on leaf cover estimates and their performance in yield prediction models

Field experiments were conducted in 1996 and 1997 to study the effects of crop growth stage and images recording height (size of the area included in the image) on the estimates of leaf cover obtained through digital image analysis techniques, and to test the effectiveness of these data in maize yield prediction. Average maize leaf cover estimates were highly variable and the precision was influenced by both the crop growth stage and the camera recording height. Sampling images from 3.3 m above the ground (1.63 m 2/image) produced less variable estimates than when sampling was done at lower heights (smaller area per image). Early timing of leaf cover assessment produced estimates with smaller variability than later timings. Maize yield prediction was slightly affected by the timing of leaf cover sampling. However, sampling at the six-leaf stage of maize may be more appropriate from a practical stand point than earlier or later samplings. While being compatible with many control options, this would allow enough weed seeds to germinate. Increasing the images recording height and consequently the area included in the images improved the accuracy of yield predictions. Data from images taken at 3.3 m fitted the model better than those from images recorded at 1.5 and 1.9 m. These results indicate that appropriate timing of leaf cover assessment and appropriate selection of image shooting height may help improve the accuracy of crop yield prediction, and thereby, reduce the risk of making wrong weed control decisions.

Measurements of Contact Angles between an Oil–Water Interface and a Fiber by the ACDPAC Technique

The ACRPAC (Analysis of Capillary Rise Profile Around a Cylinder) method was modified and extended to measure the contact angles between an oil–water interface and a fiber. Basically, the accurate image of the partial capillary depression profile was acquired and digitized by applying computer digital image processing and analysis techniques. The contact angle was determined by finding the best fit of the theoretically predicted profile, i.e., the curve representing a solution of the Laplace equation of capillarity, to the physically observed liquid–liquid interface. This analysis of the capillary depression profile around a cylinder (ACDPAC) technique was used to measure the contact angles of different water–oil interfaces on cylindrical glass fibers pre-coated with FC725. The wettability of the fiber–water–oil systems with varying oil and aqueous phases was examined. In particular, the wetting effects of a cationic surfactant cetyltrimethylammonium bromide (CTAB) and an anionic surfactant sodium dodecyl sulfate (SDS) dissolved in the aqueous phase were studied by using the ACDPAC technique. The oil phases tested were two dimethyl siloxane liquids, silicone oil A-type and silicone oil B-type.

The Study of Dynamic Slug Flow Characteristics Using Digital Image Analysis—Part I: Flow Visualization

This paper reports the application of novel, digital image analysis techniques in the study of slug flow characteristics, under dynamic conditions in two-phase gas-liquid mixtures. Water and an oil of viscosity 18 cP were used for the liquid phase and carbon dioxide was used for the gas phase. Flow in a 75-mm i.d., 10-m long acrylic pipeline system was studied. Images of slugs were recorded on video by S-VHS cameras, using an audio-visual mixer. Each image was then digitized frame-by-frame and analyzed on a SGI™ workstation. Detailed slug characteristics, including liquid film heights, slug translational velocity, mixing length, and, slug length, were obtained.

Contiguous phosphorylated and non-phosphorylated domains along axonal neurofilaments.

I have investigated the phosphorylation state of the medium molecular mass neurofilament protein (NF-M) along axonal neurofilaments. Cultured embryonic sensory neurons were treated with non-ionic detergent to cause the cytoskeletal polymers to splay apart from each other. Neurofilaments were visualized by double-label immunofluorescence microscopy and the proportion of their length that stained with various NF-M antibodies was determined using digital image analysis techniques. Monoclonal antibody RMO255, which binds to NF-M independently of phosphorylation state, stained an average of 98% of the neurofilament length. In contrast, monoclonal antibody RMO55, which binds specifically to a phosphorylated epitope on NF-M, stained some neurofilaments completely, some not at all, and some along part of their length. These partly stained neurofilaments exhibited single or multiple discrete segments of staining along their length separated by segments that were unstained. The average proportion of the neurofilament length that stained with this antibody was lowest proximally (12-22%, n=3) and increased along the axon to reach a maximum distally (58-87%, n=3). A converse pattern (77-87% proximally and 2-9% distally, n=3) was observed for neurons stained with monoclonal antibody FNP7, which binds to specifically to a non-phosphorylated epitope in both NF-M and the high molecular mass neurofilament protein, NF-H. Analysis of the staining of individual neurofilaments revealed a bimodal frequency distribution in which neurofilaments were more likely to be phosphorylated along either all or none of their length than along part of their length. These observations indicate that: (a) phosphorylated and non-phosphorylated neurofilaments can coexist side-by-side in these axons, (b) neurofilaments can be composed of single or multiple contiguous phosphorylated and non-phosphorylated epitope domains along their length, (c) the proportion of the neurofilament length that is phosphorylated at these epitopes increases along the axon in a proximal-to-distal manner, and (d) the pattern of phosphorylation is non-random, generating populations of phosphorylated and non-phosphorylated neurofilaments and discrete phosphorylated and non-phosphorylated domains along individual neurofilaments.

The combination of magnetic resonance angiography and computational fluid dynamics: a critical review.

Computational Fluid Dynamics (CFD) has proven to be a valuable technique in the study of blood flow in arteries because of its capability of obtaining quantitatively velocities and wall shear stress. However, the "bottleneck" problem limiting the application of CFD is the difficulty of constructing anatomically realistic arterial geometries. In this survey, an overview is presented of the progress over the last decade in the development of magnetic resonance angiography (MRA) techniques, medical image processing and CFD, as well as the combination of these techniques in hemodynamics research. It is demonstrated that with modern angiographic techniques such as MRA, noninvasive measurement of human angiograms becomes possible and practical. Together with digital medical image processing and analysis techniques, computational models can be constructed for the "real" human arteries without making any geometric assumptions. When allied with state-of-the-art CFD codes, velocity and wall shear stress distributions, as well as particle trajectories, can be determined in the arteries.

Image Analysis Evaluation of Aggregates for Asphalt Concrete Mixtures

The performance of asphalt concrete mixtures is influenced by the properties of the included aggregates, such as grading, shape (angularity and elongation), and texture (roughness). Complete and accurate quantification of aggregate properties is essential for understanding their influence on asphalt concrete and for selecting aggregates to produce high-quality paving mixtures. Recent developments in the use of digital image analysis techniques for quantifying aggregate morphological characteristics in asphalt concrete are summarized. Image morphological characteristics were used to quantify flatness and elongation of coarse aggregates, to estimate the proportion of natural sand in fine aggregates, and to correlate aggregate characteristics with engineering properties of asphalt concrete mixtures. Image analysis of sections also revealed information about the grading, shape, and orientation of coarse aggregates in a mixture. An overview is presented of the broad range of useful pavement engineering applications of this relatively new approach for evaluating aggregate characteristics.

Development of digital image analysis techniques for the study of velocity and void profiles in slug flow

This paper describes the development of a nonintrusive flow visualization method for the quantitative study of the dynamic velocity and void profiles across the pipe cross section in slug flow. The method utilizes novel digital image analysis and computer graphics techniques. Slug flow in a 75 mm diameter, 10 m long Plexiglass pipe is recorded on video using super-VHS cameras and an audio-visual mixer. Sixty images are obtained every second and these are digitized and analyzed on a SGI© graphics workstation. Detailed information regarding the local velocity and void distribution is obtained from the analysis and is used to generate velocity and void profiles across the pipe at different distances into the slug. Results show that gas is released into the slug in the form of pulses of bubbles resulting in the existence of large, frothy, highly aerated structures in the mixing zone. The hydrodynamic boundary layer is destroyed at the slug front but begins to redevelop in the mixing zone. It becomes fully developed at the end of the mixing zone and the one-seventh power law profile for velocity profile is applicable. At the end of the mixing zone, the gas moves towards the top of the pipe and the void fraction distribution tends to a steady profile. A process dynamic model is developed to predict the variation of the average liquid holdup within the slug.

MODELING OF NON-UNIFORM SPATIAL ARRANGEMENT OF FIBERS IN A CERAMIC MATRIX COMPOSITE

In the unidirectional fiber reinforced composites, the spatial arrangement of fibers is often non-uniform. These non-uniformities are linked to the processing conditions, and they affect the properties of the composite. In this contribution, a recently developed digital image analysis technique is used to quantify the non-uniform spatial arrangement of Nicalon fibers in a ceramic matrix composite (CMC). These quantitative data are utilized to develop a six parameter computer simulated microstructure model that is statistically equivalent to the non-uniform microstructure of the CMC. The simulated microstructure can be utilized as a RVE for the micro-mechanical modeling studies. © 1997 Acta Metallurgica Inc.

Spatial decision support systems for the management of informal settlements

Informal settlements are characterized by rapid, unstructured expansion, poorly constructed buildings, and destruction of the local ecosystem. The improvement of living conditions in informal settlements is best tackled at the local level. Effective local planning and management, however, can only be made through access to adequate geo-spatial data. This paper considers a number of promising trends in the realm of spatial information and the contribution which each of these may make to greater capability, validity and accessibility in an Interactive Decision Environment (IDE) for urban planning and management. The potential development directions reported in this article offer increased capabilities for frequent data update, new directions for modeling and visualization, and wider data accessibility and dissemination. We examine the role of Geographic Information Systems (GIS) for managing living quality improvement in informal settlements. Data capture options are then explored with emphasis on the potential of imagery derived from low-cost digital aerial imagery. Developments towards automatically generating two- and three-dimensional settlement models from this imagery using remote sensing and digital image analysis techniques are outlined. We conclude with an examination of the integration of visualization technologies in IDE and their role in the dissemination of spatial (and other) information for urban planning.

Viscoelastic properties of carbon black suspension as a flocculated percolation system

Viscoelastic properties of the disperse system consisting of carbon black particles and ethylene vinyl acetate copolymer were investigated. The effect of carbon black contents on viscoelastic properties was mainly discussed with the aid of the percolation theory. The viscoelastic properties of the molten state suspension depend on the three-dimensional flocculated structure formed by carbon black particles. However, in the solid state, the dominant factor affecting the viscoelastic properties of the composite is not the flocculated structure but the viscoelasticity of disperse medium. The expansion of the flocculated network structure formed by dispersed particles was observed morphologically by electron microscope photograph with the aid of the digital image analysis technique and the fraction of the effective bonds which belong to the infinite network was calculated. Using the percolation probability parameter determined morphologically, the modulus at small strain amplitude can be predicted quantitatively by the bond percolation theory for a lattice percolation model corresponding to purely isotropic force.

Differential regulation of type-II and type-X collagen synthesis by parathyroid hormone-related protein in chick growth-plate chondrocytes.

Parathyroid hormone-related protein is a critical autocrine regulator of endochondral ossification in the growth plate, as demonstrated by the severe disruption of growth-plate structure and function in parathyroid hormone-related protein-deficient transgenic mice. In the present study, the effects of parathyroid hormone-related protein on the synthesis of collagen mRNA and protein were studied in short-term cultures of isolated chick growth-plate chondrocytes. Parathyroid hormone-related protein selectively inhibits type-X collagen protein synthesis with no significant effect on type-II collagen protein synthesis. These effects were present in all maturationally distinct populations of chondrocytes separated by countercurrent centrifugal elutriation. In cultures of resting chondrocytes, the onset of type-X collagen expression was inhibited, while the synthesis of type-X collagen was decreased in cultures of hypertrophic chondrocytes. Synthesis of type-II and type-X collagen mRNA was examined by nonradioactive in situ hybridization with synthetic oligonucleotide cDNA probes, and the level of expression was quantified using digital image analysis. Dose-dependent suppression of type-X collagen gene expression by parathyroid hormone-related protein was observed, with no significant effect on type-II collagen mRNA detected. The results were confirmed by analysis of Northern blots of total chondrocyte mRNA. These experiments demonstrated differential transcriptional regulation of type-II and type-X collagen, with selective suppression of type-X collagen expression, by parathyroid hormone-related protein in growth-plate chondrocytes. In addition, excellent agreement was found between traditional protein and mRNA analyses and microscopic digital image analysis techniques, supporting the use of this convenient and sensitive assay method. Parathyroid hormone-related protein inhibits chondrocyte maturation and is known to stimulate proliferation, suggesting that this autocrine factor may function to regulate premature hypertrophy in the growth plate.

Assessment of erythema and skin perfusion by digital image analysis and scanner laser Doppler.

In the present study, a model for measuring the intensity of skin erythema by computer-assisted digital image analysis (DIA) was elaborated. The DIA technique was compared with laser Doppler perfusion imaging (PIM) to evaluate the relationship between erythema and skin perfusion. To produce erythema, three different types of nicotine patches (Nicotinell®, Nicorette® and Nicorette® placebo) were randomly applied ventrally on the upper arms for the time period recommended by the manufactures. After removal of the patches, colour photographs for subsequent DIA were taken and laser Doppler scanning for PIM was performed. The relationship between PIM and DIA was estimated intra- and inter-individually. A statistically significant linear correlation between PIM and DIA data was found, but the relationship was relatively weak. This may be explained by the fact that DIA and PIM are influenced by different physiological factors such as skin chromophores and velocity of moving cells. The results strengthen the theory that erythema and perfusion measurements reflect separate physiological phenomena and that the methods should be considered complements to each other in the experimental situation.

AUTOMATED MEASUREMENT OF AGGREGATE INDICES OF SHAPE

ABSTRACT Aggregate angularity is a significant factor in determining pavement resistance to rutting. Determination of aggregate angularity by direct means allows a more meaningful study of the mechanisms of pavement deformation. However, direct techniques have not been developed sufficiently to allow a more advanced study of angularity effects on pavement. Utilizing modern digital imaging hardware and image analysis techniques, we report here the development of an automated system for measuring aggregate angularity. Combining high resolution video, image capture hardware, sample transport, and computerized analysis, an instrument was constructed and used to compare various aggregate samples. Differentiation of known good quality aggregate from poorer grades was possible. Shape indices were developed from basic linear measurements, and the Hough Transform technique. These indices appear to have value in quantifying aggregate shape, and in studying the mechanisms of pavement deformation, and possibly the prediction of pavement longevity.

Study of sediment distribution in the area of the Panuco river plume by means of remote sensing

The zone of Tampico-Altamira to the north-east of Mexico has undergone considerable industrial development in the last 10 years. A great number of industries discharge their waste into the Panuco River and the sea, contaminating and altering the ecological balance of the area. Employing digital image analysis techniques and field data, two dates of thematic images in which changes in the spatial distribution of suspended solids from industrial discharges, were prepared. From the statistical properties of the defined classes, the best principal components were selected and were used as input to an Isodata clustering algorithm. Such components were derived from the band groups of the visible and the infrared. The images employed are from Landsat-5 (TM): of October 1990, and October 1992. The resulting classification was edited in such a way as to include the original image in the part corresponding to the mainland, and keeping the segmentation in sediment levels only in the coastal and lacunar zone. The images were georeferenced in UTM coordinates. The results show a complex system of hydrodynamic interaction between the Panuco River, the sea and the coastal lagoons.

Effects of Processing Conditions on Birefringence Development in Injection Molded Parts. II. Experimental Measurement

The birefringence of injection molded parts was measured using a digital photoelasticity system, which combines a digital image analysis technique and the half-fringe photoelasticity (HFP) method The effects of processing conditions, including melt temperature, mold temperature, filling time and packing pressure, on the birefringence development in the molded parts were investigated. It was found that temperature and pressure are the two dominant factors that determine the birefringence development in the parts during the molding process. Frozen-in birefringence of the molded parts decreases with increasing melt temperature, mold temperature and injection speed. Birefringence of the parts also increases with increased packing pressure, especially around the gate area. Numerical simulations using the Leonov viscoelastic fluid model predict similar dependence of birefringence of parts on processing conditions. Simulated results are also consistent with measured values.

Viability Analysis of Cryopreserved Rat Pancreatic Islets Using Laser Scanning Confocal Microscopy

We have developed a digital image analysis technique to assay the viability of frozen–thawed pancreatic islets by using laser scanning confocal microscopy (LSCM) in conjunction with double fluorescent staining [acridine orange/propidium iodide (AO/PI)]. Freshly isolated rat pancreatic islets were cultured for 18–24 h and then brought to a 2Mconcentration of dimethyl sulfoxide (Me2SO) by serial addition at decreasing temperatures. Ice was nucleated in the islet suspension at a defined temperature (−10°C), followed by a controlled period for equilibration and then cooling in a programmable bulk freezer at rates of 0.3, 1, 3, 10, and 30°C/min to −70°C. Samples were then stored in liquid nitrogen. Subsequent to rapid thawing and serial dilution with sucrose solution to remove Me2SO, AO/PI-stained individual islets were prepared for imaging on the LSCM. A series of optical sections through individual stained islets were obtained and processed to obtain high-contrast images at two different wavelengths; 488 nm and 514 nm for viable and damaged tissue, respectively. Image analysis algorithms consisted of template masking, generation of histograms of the pixel intensity profile, and gray level thresholding to obtain binary images. The total percentages of both types of tissue in the islet were computed by summing the two populations in each serial section. The spatial distributions of viable and damaged tissue were calculated from the three-dimensional (3-D) data base for both cultured (control) and cryopreserved islets. The 3-D spatial distributions of damaged and viable tissue in the islets were computed by determining the normalized distance of each viable/damaged voxel from the centroid of the islet volume to a mathematically estimated 3-D superquadric surface used to estimate the outer boundary of the islet. Further, each isolated damaged cell was identified and its volume determined. These studies indicate that cryopreserved islets exhibit shape distortion and a decrease in the numerical density of cells in comparison to unfrozen controls. Maximal survival was observed at the slower cooling rates. Accordingly, damage was found to occur throughout the 3-D islet volume in distinct spatial distributions for islets frozen at the slower and the faster cooling rates. Further, it was found that the volume of the majority of damaged cells identified was consistent with that of cells ranging in diameter from 5 to 9 μm.

Verification of the microdomain model of crystallinity in polyvinyl chloride by electron microscopy

A study of the fine scale microstructure of PVC was carried out using a combination of high resolution transmission electron microscopy and digital image analysis techniques. The images obtained contained a degree of order of the approximate size and distribution as predicted by the microdomain model of crystallinity in PVC. The microdomain model for crystallinity in PVC has been built up from previous studies using various analytical techniques including wide- and small-angle x-ray diffraction and differential scanning calorimetry. Earlier studies using transmission electron microscopy did not find any direct evidence supporting this model. Significant advances in both electron microscope and image processing technology had taken place since the earlier microscopic studies. The TEM imaging and image analysis procedures that have been utilized in this research may be applicable to the imaging of very fine scale ordering in other polymers.