Perceptual primitives from an extended 4D Hough transform

Directional features extracted from Gabor wavelets responses were used to train a structure of self-organising maps, thus classifying each pixel in the image within a neuron-map. Resulting directional primitives were grouped into perceptual primitives introducing an extended 4D Hough transform to group pixels with similar directional features. These can then be used as perceptual primitives to detect salient structures. The proposed method has independently fixed parameters that do not need to be tuned for different kind or quality of images. We present results in application to noisy FLIR images and show that line primitives for complex structures, such as bridges, or simple structures, such as runways, can be found by this approach. We compare and demonstrate the quality of our results with those obtained through a parameter-dependent traditional Canny edge detector and Hough line finding process.     

TEXEMS: texture exemplars for defect detection on random textured surfaces

We present an approach to detecting and localizing defects in random color textures which requires only a few defect free samples for unsupervised training. It is assumed that each image is generated by a superposition of various-size image patches with added variations at each pixel position. These image patches and their corresponding variances are referred to here as textural exemplars or texems. Mixture models are applied to obtain the texems using multiscale analysis to reduce the computational costs. Novelty detection on color texture surfaces is performed by examining the same-source similarity based on the data likelihood in multiscale, followed by logical processes to combine the defect candidates to localize defects. The proposed method is compared against a Gabor filter bank-based novelty detection method. Also, we compare different texem generalization schemes for defect detection in terms of accuracy and efficiency.     

Designing and planning a multi-echelon multi-period multi-product closed-loop supply chain utilizing genetic algorithm

Designing and planning a closed-loop supply chain in a comprehensive structure is vital for its applicability. To cope with the design and planning issue of a comprehensive closed-loop supply chain network, this paper develops an extended model, which is multi-echelon, multi-product, and multi-period in a mixed integer linear programming framework. The word “comprehensive,” in our mathematical approach, in designing and planning a closed-loop supply chain problem, can be analyzed from two complementary angles: including all possible entities (facilities) of a real condition and considering minimum limitations on possible flows between entities. In our proposed model, customers can be supplied via manufacturers, warehouses, and distributors, as an example. The proposed model is solved by CPLEX optimization software and by a developed genetic algorithm. During this computational analysis, we compare results of proposed pretuned genetic algorithm with a global optimum of CPLEX solver. Then, a sufficient number of large-size instances are generated and solved by the proposed genetic algorithm. To the best of our knowledge, there has been no similar multi-period multi-product closed-loop supply chain design and planning problem utilizing any kind of meta-heuristics let alone genetic algorithms. Therefore, in this issue, it is an original research, and results prove the acceptable performances of the developed genetic algorithm.     

Incorporating risk measures in closed-loop supply chain network design

This paper considers a location-allocation problem in a closed-loop supply chain (CLSC) with two extensions: first, demand and prices of new and return products are regarded as non-deterministic parameters and second, the objective function is developed from expected profit to three types of mean-risk ones. Indeed, design and planning an integrated CLSC in real-world volatile markets is an important and necessary issue. Further, risk-neutral approaches, which are considered expected values, are not efficient for such uncertain conditions. Hence, this paper, copes with the design and planning problem of a CLSC in a two-stage stochastic structure. Besides, risk criteria are considered through using three types of popular and well-behaved risk measures: mean absolute deviation, value at risk and conditional value at risk (CVaR). Consequently, three types of mean-risk models are developed as objective functions and decision-making procedures are undertaken based on the expected values and risk adversity criteria. Finally, performances of the developed mean-risk models are evaluated in various aspects. Results reveal that the inefficiencies of risk-neutral approaches can be overcome. In addition, in terms of quality of solutions, the acceptability of CVaR is proved when it is compared to other risk measures.     

Spraying Cellulose Nanofibrils for Improvement of Tensile and Barrier Properties of Writing & Printing (W&P) Paper

An experimental spray coater was used to coat writing and printing (W&P) paper substrate with cellulose nanofibrils (CNFs) suspensions. The effects of spraying variables (i.e. concentration of suspension, spray pressure, distance and time of spray) on the coated sheets were analyzed in terms of the tensile strength, water vapor transmission rate (WVTR), and oxygen transmission rate (OTR). Basis weight and the thickness of coated layers in the different treatments were measured. In addition, image analysis of the microstructure examined the coating adhesion. The WVTR of the papers decreased, while tensile strength increased with one layer of CNF coating. The OTR was not changed with the CNF coating. The tensile strength and microstructure images of the coated papers indicate good adhesion between the CNF coating and the paper substrate when using the spray coater.     

Feedback control strategies for object recognition

We present a paradigm for feedback strategies that find instances of a generic class of objects by improving on established single-pass hypothesis generation and verification approaches. We improve upon the mechanisms of the traditional or classical image processing systems by introducing control strategies at low, intermediate, and high levels of analysis. We produce optimal sets of low-level features to reduce the number of hypotheses generated. The feedback further enables updated sets of features to be extracted so that the target object may be located even in very, noisy data. The use of an interest operator in the feedback directs the search through the hypotheses in an optimal manner, so minimizing the amount of feedback to false alarms. Furthermore, we aim to obtain detailed information about a complex object and not just its location. Thus, following top-down recognition of the object our feedback control directs the search for missing information. The system can extract complex objects in a scale and rotation independent manner where the objects may be partially occluded. The method is illustrated using box shaped objects and noisy IR images of a number of bridges     

Recognising text in real scenes

We present two different approaches to the location and recovery of text in images of real scenes. The techniques we describe are invariant to the scale and 3D orientation of the text, and allow recovery of text in cluttered scenes. The first approach uses page edges and other rectangular boundaries around text to locate a surface containing text, and to recover a fronto-parallel view. This is performed using line detection, perceptual grouping, and comparison of potential text regions using a confidence measure. The second approach uses low-level texture measures with a neural network classifier to locate regions of text in an image. Then we recover a fronto-parallel view of each located paragraph of text by separating the individual lines of text and determining the vanishing points of the text plane. We illustrate our results using a number of images. Received May 20, 2001 / Accepted June 19, 2001     

Segmentation of color textures

This paper describes an approach to perceptual segmentation of color image textures. A multiscale representation of the texture image, generated by a multiband smoothing algorithm based on human psychophysical measurements of color appearance is used as the input. Initial segmentation is achieved by applying a clustering algorithm to the image at the coarsest level of smoothing. The segmented clusters are then restructured in order to isolate core clusters, i.e., patches in which the pixels are definitely associated with the same region. The image pixels representing the core clusters are used to form 3D color histograms which are then used for probabilistic assignment of all other pixels to the core clusters to form larger clusters and categorise the rest of the image. The process of setting up color histograms and probabilistic reassignment of the pixels to the clusters is then propagated through finer levels of smoothing until a full segmentation is achieved at the highest level of resolution     

QUAC: Quick unsupervised anisotropic clustering

We present a novel unsupervised algorithm for quickly finding clusters in multi-dimensional data. It does not make the assumption of isotropy, instead taking full advantage of the anisotropic Gaussian kernel, to adapt to local data shape and scale. We employ some little-used properties of the multivariate Gaussian distribution to represent the data, and also give, as a corollary of the theory we formulate, a simple yet principled means of preventing singularities in Gaussian models. The efficacy and robustness of the proposed method are demonstrated on both real and artificial data, providing qualitative and quantitative results, and comparing against the well known mean-shift and K-means algorithms.     

Predicting fracture of solder joints with different constraint factors

Double cantilever beam (DCB) specimens of 2.5‐mm‐long SAC305 solder joints were prepared with thickness of copper adherends varying from 8 to 21 mm each. The specimens were tested under mode I loading conditions (ie, pure opening mode with no shear component of loading) with a strain rate of 0.03 second^?1. The measured fracture load was used to calculate the critical strain energy release rate for crack initiation, J_ci, in each case. Fracture behaviour showed a significant dependence on the adherend thickness; the J_ci and plastic deformation of the solder at crack initiation decreased significantly with increase in adherend thickness. This behaviour was attributed to changes in stress distribution along the solder layer when the adherend thickness was varied. The capability of J_ci as a property was then assessed to predict the fracture load of solder joints in specimens with different constraint levels caused by variations in adherend thicknesses. In light of the results obtained, a cohesive zone model (CZM) was developed to predict the fracture load of solder joints as a function of adherend thickness. Finally, a CZM with a single set of parameters was established to predict the fracture loads for all the cases. It was concluded that CZM was a better methodology to account for changes in degree of joint constraint imposed by bonding adherends.