Sequence Of Image Processing Algorithms Research Articles

Defect Detection of Coatings on Metal Surfaces Based on K-Means Clustering Algorithm

Defect detection is an important phase for the analysis of the surface quality of products as it influences the subsequent process. The existence of surface defects will affect the corrosion and wear resistance of the end product. The insufficient detection and classification rates of the standard algorithms are infeasible to accomplish the production requirements. Surface defect detection in coated metals with non-destructive techniques is an essential prerequisite for quality analysis in manufacturing stage. The existence of surface defects can significantly alter the deterioration resistance and instinctive qualities of a material and as a result more expansive analysis is essential. This paper proposes a competent and exact approach using K-means algorithm for the detection of surface coating defects. K-means is an unsupervised algorithm used for segmenting the area of interest from background. The proposed method uses a sequence of image processing algorithms to examine and validate the input image real-time accuracy for detection of defects. The proposed method efficiency is featured with test samples and results from experimental analysis. It shows that the proposed method be able to adequately and instinctively recognize the presence of defects inside coatings on metal surfaces.

Characteristics of sliding bubbles in subcooled flow boiling in a narrow rectangular channel under natural circulation condition

Narrow rectangular channels are widely used in many fields due to its characteristics of boiling heat transfer enhancement, and the bubble behaviors have been in particular interest for decades because of its significant contribution to understanding the mechanism of heat transfer enhancement. In the present work, visualization experiments in upward subcooled flow boiling in a narrow rectangular channel under natural circulation condition were carried out. A sequence of image processing algorithms was used to deal with the original bubble images to get relevant bubble parameters, including bubble diameter and bubble velocity, etc. The experiments were performed at pressures ranging from 0.1 MPa to 0.3 MPa, with inlet subcooling ranging from 20 to 60 K and heat flux ranging from 100 kW/m2 to 300 kW/m2. Effects of the liquid subcooling and heat flux on the bubble stochastics and mean characteristics were investigated. In natural circulation, the bubble diameter distribution approximately conforms to lognormal distribution while the bubble velocity distribution agrees well with normal distribution. These are in contrast to both of the bubble diameter and velocity follow normal distribution in forced circulation. The present experiment data shows a good linear relationship between the mean bubble velocity and mean bubble diameter. A correlation for the mean bubble diameter as a function of local subcooling, heat flux, mass flux, and gap width size were obtained. Moreover, we found that all the bubbles will slide along the heating wall after generation in narrow rectangular channel, and two types of sliding bubbles can be observed in our experiments. The growth curve of the Lift-type bubbles and the Sliding-type bubbles conform to a dual model and a power model, respectively.

Visualized Experiment of Bubble Behaviors in a Vertical Narrow Rectangular Channel Under Natural Circulation Condition

The characteristics of bubble behavior have been in particular interest for decades due to its significant contribution to understanding the mechanism of heat transfer. In the present work, visualized experiment is conducted to study the bubble characteristics in subcooled flow boiling of a narrow rectangular channel under natural circulation. The experiments were performed at pressures of 0.2 MPa, with inlet subcooling ranging from 20 to 60 K and heat flux ranging from 100 kW/m2 to 300 kW/m2. A high-speed digital camera is used to capture the pictures of bubble behaviors. A sequence of image processing algorithms is used deal with the original bubble images to get relevant bubble parameters. We observe the whole process of a single sliding bubble lifetime and found most of bubbles slide along the heating surface after detaching the nucleation sites. Five typical sliding bubble growth paths are observed in the present experimental conditions. According to the analysis of the experimental data, it can be found that the liquid subcooling and wall superheat are the main factors that affect the bubble size during sliding in narrow rectangular channel under natural circulation condition. Due to the difference of driving force, the sliding velocity of bubble in forced circulation is always larger than that in natural circulation. At the same time, the bubble velocity changes significantly at different heat flux and shooting location.

Automatic localization of landmark sets in head CT images with regression forests for image registration initialization.

Cochlear Implants (CIs) are electrode arrays that are surgically inserted into the cochlea. Individual contacts stimulate frequency-mapped nerve endings thus replacing the natural electro-mechanical transduction mechanism. CIs are programmed post-operatively by audiologists but this is currently done using behavioral tests without imaging information that permits relating electrode position to inner ear anatomy. We have recently developed a series of image processing steps that permit the segmentation of the inner ear anatomy and the localization of individual contacts. We have proposed a new programming strategy that uses this information and we have shown in a study with 68 participants that 78% of long term recipients preferred the programming parameters determined with this new strategy. A limiting factor to the large scale evaluation and deployment of our technique is the amount of user interaction still required in some of the steps used in our sequence of image processing algorithms. One such step is the rough registration of an atlas to target volumes prior to the use of automated intensity-based algorithms when the target volumes have very different fields of view and orientations. In this paper we propose a solution to this problem. It relies on a random forest-based approach to automatically localize a series of landmarks. Our results obtained from 83 images with 132 registration tasks show that automatic initialization of an intensity-based algorithm proves to be a reliable technique to replace the manual step.

Color grading of beef fat by using computer vision and support vector machine

Machine vision and support vector machine (SVM) were used to determine color scores of beef fat. One hundred and twenty-three of beef rib eye steaks were selected to sensory evaluation and image processing. After fat color score was assigned to each steak by a five-member panel according to the standard color cards, images were acquired for each steak. The subcutaneous fat was separated from the rib eye by using a sequence of image processing algorithms, boundary tracking, thresholding and morphological operation, etc. Twelve features of fat color (six features were extracted from the subcutaneous fat images and the other six were calculated) were used as input for SVM classifiers. The best SVM classifier was chosen according to percentage of correct classified samples based on the training set and then was validated by a nondependent test set. The proposed SVM classifier achieved the best performance percentage of 97.4%, showing that the machine vision combined with SVM discrimination method can provide an effective tool for predicting color scores of beef fat.

Pizza sauce spread classification using colour vision and support vector machines

An automated classification system of pizza sauce spread using colour vision and support vector machine (SVM) was developed. To characterise pizza sauce spread with low dimensional colour features, a sequence of image processing algorithms was developed. After image segmentation from the background, the segmented image was transformed from red, green, and blue (RGB) colour space to hue, saturation, and value (HSV) colour space. Then a vector quantifier was designed to quantify the HS (hue and saturation) space to 256-dimension, and the quantified colour features of pizza sauce spread were represented by colour histogram. Finally, principal component analysis (PCA) was applied to reduce the 256-dimensional vectors to 30-dimensional vectors. With the 30-dimensional vectors as the input, SVM classifiers were used for classification of pizza sauce spread. It was found that the polynomial SVM classifiers resulted in the best classification accuracy with 96.67% on the test experiments.

Automated extraction of coastline from satellite imagery by integrating Canny edge detection and locally adaptive thresholding methods

This paper presents a comprehensive approach to effectively and accurately extract coastlines from satellite imagery. It consists of a sequence of image processing algorithms, in which the key component is image segmentation based on a locally adaptive thresholding technique. Several technical innovations have been made to improve the accuracy and efficiency for determining the land/water boundaries. The use of the Levenberg-Marquardt method and the Canny edge detector speeds up the convergence of iterative Gaussian curve fitting process and improves the accuracy of the bimodal Gaussian parameters. The result is increased reliability of local thresholds for image segmentation. A series of further image processing steps are applied to the segmented images. Particularly, grouping and labelling contiguous image regions into individual image objects enables us to utilize heuristic human knowledge about the size and continuity of the land and ocean masses to discriminate the true coastline from other object boundaries. The final product of our processing chain is a vector-based line coverage of the coastline, which can be readily incorporated into a GIS database. Our method has been applied to both radar and optical satellite images, and the positional precision of the resulting coastline is measured at the pixel level.

Impaired cerebral cortical gray matter growth after treatment with dexamethasone for neonatal chronic lung disease.

The specific aim of this study was to quantify at term the influence of postnatal systemic dexamethasone treatment for neonatal chronic lung disease on subsequent brain growth and development in premature infants without evidence of severe intraventricular hemorrhage or white matter injury. Eighteen premature (23 to 31 weeks) infants, 7 treated with dexamethasone and 11 not treated, were studied at term, ie, 38 to 41 postconceptional weeks, by an advanced quantitative volumetric 3-dimensional magnetic resonance imaging (MRI) technique to quantify cerebral tissue volumes. Fourteen healthy term infants also were studied for comparison. A sequence of image processing algorithms was used to segment each of the MRI slices into the following separate tissue classes: cerebral cortical gray matter, basal ganglia/thalami, unmyelinated white matter, myelinated white matter, and cerebrospinal fluid, all classified based on magnetic resonance signal intensity and anatomic location. A final summing of voxels for each tissue class was performed to compute absolute volumes in milliliters. Cerebral cortical gray matter volume in premature infants treated with dexamethasone was reduced 35% when compared with gray matter volume in premature infants not treated with dexamethasone (mean +/- standard deviation, 130.3 +/- 54.0 vs 200.6 +/- 35.1 mL, respectively). Subcortical gray matter volumes (basal ganglia and thalami) and myelinated and unmyelinated white matter volumes were not significantly different among the treated and untreated groups. However, premature infants treated with dexamethasone exhibited a reduction (30%) in total cerebral tissue volume compared with total cerebral tissue volume in both the premature infants not treated with dexamethasone and the control term infants (312.7 +/- 43.7 vs 448.2 +/- 50.2 and 471.6 +/- 36.4 mL respectively). This latter finding relates primarily to the decrease in cerebral cortical gray matter volume. The data suggest an impairment in brain growth, principally affecting cerebral cortical gray matter, secondary to systemic dexamethasone therapy. Although the premature infants who received dexamethasone were smaller with more severe respiratory disease, these findings are consistent with growing evidence of a potential deleterious effect of dexamethasone on neonatal brain and subsequent neurodevelopmental outcome. This apparent deleterious effect should be taken into consideration by clinicians when weighing the potential risks and benefits of this therapy for low birth weight infants with neonatal chronic lung disease.

Image processing algorithms for the analysis of phase-shifted speckle interference patterns

A sequence of image processing algorithms for the analysis of interference patterns generated by a phaseshifting speckle interferometer is discussed. The goal is the accurate determination of displacement and strain components at the surface of an object. The phase change related to the displacement is accurately calculated from eight digitized interference patterns using a phase-shifting algorithm. Digital image processing algorithms have been developed for phase unwrapping, phase restoration, and phase fitting. During the processing steps a binary mask is used to solve the problem of invalid areas. Experimental results for the strain components at the surface of a simple object demonstrate a repeatability of 0.3-microstrain rms.