Local Gray Level Research Articles

Semi-blind image restoration using a local neural approach

This work aims to define and experimentally evaluate an iterative strategy based on neural learning for semi-blind image restoration in the presence of blur and noise. Salient aspects of the proposed strategy are the use of a local error function derived from the conventional global constrained error measure and the assignment of a separate regularization parameter to each image pixel based on local gray level variance. This method can be viewed as a neural strategy where the pixels of the restored image are the synapse's weights the neural network tries to modify during learning to minimize the output error measurement. The method was experimentally evaluated in terms of restoration quality and speed using test images synthetically degraded and increasingly corrupted. To investigate whether the strategy can be considered an alternative to neural restoration procedures, the results were compared with those obtained by well known Hopfield-based restoration approaches. Results obtained show that our method performs significantly better and faster than other models considered.

Volume quantification by fuzzy logic modelling in freehand ultrasound imaging

Introduction Many algorithms exist for 3D reconstruction of data from freehand 2D ultrasound slices. These methods are based on interpolation techniques to fill the voxels from the pixels. For quantification purposes, segmentation is involved to delineate the structure of interest. However, speckle and partial volume effect errors can affect quantification. Objective This study aimed to assess the effect of the combination of a fuzzy model and 3D reconstruction algorithms of freehand ultrasound images on these errors. Methods We introduced a fuzzification step to correct the initial segmentation, by weighting the pixels by a distribution function, taking into account the local gray levels, the orientation of the local gradient, and the local contrast-to-noise ratio. We then used two of the most wide-spread reconstruction algorithms (pixel nearest neighbour (PNN) and voxel nearest neighbour (VNN)) to interpolate and create the volume of the structure. Finally, defuzzification was used to estimate the optimal volume. Validation B-scans were acquired using 5 MHz and 8 MHz ultrasound probes on ultrasound tissue-mimicking phantoms. Quantitative evaluation of the reconstructed structures was done by comparing the method output to the real volumes. Comparison was also done with classical PNN and VNN algorithms. Results With the fuzzy model quantification errors were less than 4.3%, whereas with classical algorithms, errors were larger (10.3% using PNN, 17.2% using VNN). Furthermore, for very small structures (0.5 cm 3), errors reached 24.3% using the classical VNN algorithm, while they were about 9.6% with the fuzzy VNN model. Conclusion These experiments prove that the fuzzy model allows volumes to be determined with better accuracy and reproducibility, especially for small structures (<3 cm 3).

Quantitative evaluation of local myocardial blood volume in contrast echocardiography

Myocardial perfusion is usually assessed by Single Photon Emission Computed Tomography (SPECT) imaging. Information about myocardial perfusion is sometimes deduced from angiography or Computed Tomography (CT) angiography, which detect coronary artery stenosis. Contrast echocardiography can be used for that purpose as well. However, the currently available data acquisition and analysis methods are difficult to manage in the clinical environment. This paper presents a novel contrast echo data acquisition protocol and parameter extraction procedure, providing an automatic quantitative evaluation of the local myocardial blood volume for the entire left ventricular myocardium. This information is indicative of local perfusion. Our method evaluates the myocardial blood volume according to the local gray level intensity, as measured during a single heartbeat, when there is a distinct myocardial opacification (based on visual estimation). The echocardiographic image analysis is based on a new attenuation correction technique, which compensates for the ultrasonic signal attenuation in both the tissue and the contrast agent. In comparison, the existing contrast echo based methods utilize the long-term temporal variability of the gray level to extract information regarding the local myocardial blood flow velocity. Our technique has been tested on 17 cine-loops of 15 different patients. We have found a high correlation between abnormal segments, detected automatically by our technique, and segments that have been clinically diagnosed as ischemic (at rest) or infarcted. For that purpose, we have defined ischemic segments as segments fed by coronary arteries with severe stenosis, as determined by angiography, and infarcted segments as segments after Acute Myocardial Infarction, as detected by electrocardiography. Furthermore, we have found a high correlation between the automatically calculated myocardial blood volume levels and the clinical evaluation of segmental contractility, based on echocardiographic imaging.

Dedicated Image Analysis Software Tool for the Evaluation of the Resorption Activity of Cultured Osteoclasts

A semi-automated image analyzing system was developed for the quantitative study of the in vitro osteoclast activity. It is able to determine the percentage of the matrix reabsorbed by cultured osteoclasts. It counts and sizes the worn areas and displays the results together with two-color coding useful for visual interpretation. A specific semi-automated procedure was designed to simulate human sight. It detects local gray level differences while matching with the surrounding gray levels. It implements routine calculation enabling the operator to continuously control the process. We were able to accurately and precisely quantify the reabsorbed percentage of meaningful fields in a slice. This image-analyzing system is useful in the study of physiological or pathological osteoclast activity, in various diseases involving bone, such as osteoporosis, Paget's disease of bone or neoplastic conditions, and in the evaluation of osteoclast response to various agents.

Local measurement of oxygen transfer around a single bubble by planar laser-induced fluorescence

The aim of this work is to develop a non-intrusive experimental technique to measure oxygen concentration in a liquid phase. This technique relies on the planar laser-induced fluorescence (PLIF). In its classical version, PLIF consists in viewing the presence of a fluorescent dye in a liquid phase by means of a laser sheet. Digital image analysis gives the relationship between local grey levels and the dye concentration. The principle of the oxygen concentration measurement rests on the fact that the oxygen molecules inhibit this fluorescence, in proportion to their concentration. After calibration, analysis of the grey levels gives an image of the 2D field of oxygen concentration.

Quantifying image quality at breast periphery vs mammary gland in mammography using wavelet analysis.

Use of high contrast film-screen systems in mammography, in combination with the fact that exposure parameters are selected to ensure good visualization of the mammary gland, results in overexposure of the film area corresponding to the breast periphery, therefore decreasing image quality. The aim of this work was to provide a quantitative evaluation of image quality at the breast periphery compared with the mammary gland. To deal with the difficulties in quantification of image quality introduced by low contrast encountered at the breast periphery, wavelet analysis has been used for derivation of a contrast indicator (CI) and a noise indicator (NI), taking into account local grey level variations. Gradient magnitude coefficients corresponding to region of interest (ROI) grey level values are the basis of CI definition. Mammary gland and breast periphery were sampled by equally spaced ROIs, the quantity of which was determined by a heuristic method. For NI definition, the power values of gradient magnitude coefficients corresponding to the ROI were utilized. Image quality at the breast periphery compared with the mammary gland was evaluated using 150 craniocaudal images from the Digital Database for Screening Mammography. Measurements were carried out using a tool developed in our department. A 50% contrast decrease at the breast periphery was observed, while noise decreased by approximately 2%.

Realizations of fast 2-D/3-D image filtering and enhancement.

This paper proposes a novel algorithm for multidimensional image enhancement based on a fuzzy domain enhancement method, and an implementation of a recursive and separable low-pass filter. Considering a smoothed image as a fuzzy data set, each pixel in an image is processed independently, using fuzzy domain transformation and enhancement of both the dynamic range and the local gray level variations. The algorithm has the advantages of being fast and adaptive, so it can be used in real-time image processing applications and for multidimensional data with low computational cost. It also has the ability to reduce noise and unwanted background that may affect the visualization quality of two-dimensional (2-D)/three-dimensional (3-D) data. Examples for the applications of the algorithm are given for mammograms, ultrasound 3-D images, and photographic images.

A novel method for quantifying shape deformation applied to biocompatibility testing.

Cytotoxicity tests are important for the screening and evaluation of biocompatibility of artificial organs. Morphologic changes of cells that were contacted biomaterials or biomaterial extracts indicate their toxicity. However, information on cytotoxic effects is still obtained by subjective visual inspection of microscopic samples. In this article, a novel computer assisted method is introduced. The automatic analysis of digitized micrographs is achieved in several stages: segmentation, separation, classification, and measurement. The segmentation of the image is provided by a new local adaptive thresholding technique, which adapts the threshold window sizes onto local gray level distribution and yields optimal window sizes. The actual threshold is obtained by maximizing interclass variances and minimizing intraclass variance. For the separation of connected cells, the binarized samples are cleaned from "false" markers by morphologic filtering. The subsequent separation is a two phase approach. Information levels are generated top-down by successively applying an enhanced erosion operator, which yields markers and filters noise usually evolving from multiple erosions. The converse bottom-up integration of the eroded markers is performed by successively applying an enhanced dilation operator, which reconstructs the cells and prevents merging of already separated objects. The subsequent measuring provides quantitative parameters of the distribution of size and compactness of the cells contained within the sample. The method was evaluated by L-929 fibroblasts that were in contact with 0%, 5%, and 10% concentrations of ethanol. For each concentration, 268 images of the cell populations were captured. The obtained quantitative parameters are highly correlated to the common verbal description of morphologic changes. Therefore, the proposed automatic method has several advantages compared with subjective examinations. The results allow an objective comparison of the quantification of phenomena; the subjective influence of the observer is eliminated; and the laboratory staff is relieved of time consuming routine work.

Image Contrast Enhancement by Constrained Local Histogram Equalization

Histogram equalization is a widely used image contrast enhancement method. While global histogram equalization enhances the contrast of the whole image, local histogram equalization can enhance many image details by taking different transformation of the same gray level at different places in the original image. However, the local histogram equalization process often results in unacceptable modification of the original image appearance. In this paper, a constrained local histogram equalization method is proposed to balance the conflicting requirements: enhancement of the image details and the maintenance of the overall image appearance. Our method uses the variational form of histogram equalization so that a constraint condition, which forces the local gray level transformations to change continuously in the spatial domain, can be introduced into the equalization process. Experimental results of different kinds of images show the effect of our method.

The application of local grey level histograms to organelle classification in histological images

The performance of a histogram matching algorithm was evaluated for the classification of organelles in histological images. The test set used for this investigation comprised several digital images of sheep pituitary gland captured using a CCD camera linked to a Tunnelling Electron Microscope. Results show that the normalized image histogram is a good feature to choose as the basis of object recognition in biological images.

Automated image analyzing system for the quantitative study of living cells in culture

A fully automated image analyzing system was developed for the quantitative study of cells in culture. It was able to count cells, to classify cells according to their morphological characteristics and to follow cell culture development. A specific procedure was designed to process Hoffman modulation contrast images. It detects local gray level differences while using conditional dilation techniques. We were able to successfully detect aggregated unstained cells, presently a technical limit in image segmentation. Living cells can be studied in a noninvasive and nondestructive way with this system. An improved automatic focusing algorithm was developed which ensured an accurate prediction of the optimal focus position. A strictly defined sampling procedure was applied to estimate unbiasedly cell density and obtain precisely cell contours. The evaluation of the system was carried out on Chinese hamster ovary (CHO-NTR) cell cultures treated with a newly developed neurotensin agonist JMV449. Chinese hamster ovary cell division was found to be retarded 20 hours after the JMV449 treatment, while the morphology of CHO-NTR cells has already undergone significant changes 12 hours after the treatment. This image analyzing system provides the possibility to follow cell culture development (e.g., cell density evolution, cell morphological changes) under various experimental conditions.

Parallel implementation of local averaging for image processing

We present an efficient image processing architecture to compute, repeatedly over the pixels of an image, the mean inside a rectangular window. Thus, the local average grey levels are obtained. The designed pixel-oriented architecture has been systematically derived using techniques for synthesizing systolic arrays from recurrence equations.

Statistical Edge Detection Operators for Linear Feature Extraction in SAR Images

Edge detection in synthetic aperture radar (SAR) images is rendered difficult by the presence of speckle. The data are often filtered using adaptive filters independently of the edge detection process when, in fact, the two steps should be coupled (i.e., the local homogeneity criterion employed by an adaptive filter should be consistent with the edge detector criterion). Five different edge detection algorithms for SAR images are evaluated and compared. The detection algorithms are comprised of two operators based on non-parametric statistical tests, the Ratio of Averages test, difference of averages (essentially a gradient method), and a test based on the mean squared to variance ratio. Two edge thinning and thresholding operations are also compared: an algorithm proposed by Nevatia and Babu (1980), and one based on mathematical morphology (Serra, 1980). Initial testing is carried out on simulated imagery for accurate control of the signal being masked by speckle noise. We obtain the best results using the ratio operator in combination with the morphological thinning operations. High real edge recovery rates are required for segmentation (Fmn > 0.95), and these levels of Fmn are only produced by high-contrast, low mean, local grey level boundaries. We show that a complete boundary delineation for segmentation purposes cannot be expected from a typical Seasat SAR agricultural scene due to the large number of low contrast edges. These methods are applicable in situations where there is a greater contrast between the targets to be discriminated. A suitable application of edge extraction from SAR, for example, is lake boundary extraction; a sample image is presented.

Symbolic pixel labeling for curvilinear feature detection

This paper describes a method of detecting thin curvilinear features in an image based on a detailed analysis of the local gray level patterns at each pixel. This allows operations such as thinning and gap filling to be based on more accurate information.

Segmentation of fingerprint images using the directional image

In this paper the concept of the directional image is introduced and its application for segmentation is presented. The directional image can be thought of as an image transform, where each pixel of the image represents direction of the local grey level uniformity. The statistics derived from the directional image are used for the segmentation of the original image. It has been found to work better than the segmentation based on grey level statistics and other methods. This method is suited for simple images like Fingerprint and other images that consist of only background and foreground, but whose histogram may not necessarily be bimodal. The results of segmentation with the new and other methods are presented.

Personal Computer Based Image Processing With Halftoning

A versatile and useful personal computer based procedure to generate halftone representation of digitized image data is presented. This feature provides limited interactive image processing capability for a basic personal computer work station without having to acquire a video monitor and a digital-to-analog converter board. The generated halftone images can be displayed on the standard dot-addressable graphics monitor, and/or they can be printed on regular paper via an associated printer. The halftoning procedure is based on an error diffusion technique in which dots are produced randomly, one at a time, based on both the local gray level value and some accumulated measure of the error due to previous assignments. The key elements in halftoning, such as dynamic, spatial, and display medium resolutions, are defined, and the mathematical correspondence among them is established.

Compairing images of image sequences using segments

A new method for the comparison of images (frames) of image sequences is proposed. Instead of image functions (e.g. correlation) symbolic image descriptions referring to segmentation results for single frames are compared. This paper describes the chosen methods for image segmentation which refer to an iterative binarization (level slicing) and to a region growing (local grey level evaluation) algorithm. The generated segments characterized by features and attributes represent a symbolic description and are the basis for the comparison of successive frames. For the matching of segments the hypotheses no changes, pseudo changes, and real changes are tested. The applied algorithms refer to the similarity of shape features, to spatial relationships and to contour points. First results are given for the image segmentation and the image comparison of test scenes.

Statistical filter for image feature extraction

Edge extraction techniques have become important as a preprocessing step in extraction of image features for the purpose of image segmentation, object identification, and bandwidth compression. The use of conventional edge extractors such as Sobel and Laplacian filters results in images that in many cases have a high degree of clutter due to the natural spatial texture of the scene background. To overcome this difficulty, a statistical filter has been developed that enhances local grey level activity around objects while reducing contributions due to background. The statistical filter is employed in a neighborhood modification process where the central pixel is replaced with the third central moment computed from the surrounding neighborhood.Choice of the third central moment is due in part to the fact that it is a function of the scene within the neighborhood rather than the power spectral density (Wiener spectrum) of the neighborhood. Application of the filter requires no prior knowledge, and pixels within the filter window may be chosen in random order due to the statistical nature of the operation. Results of the filter applied to IR images show performance comparable with, and in some cases superior to, the Sobel and Laplacian filters most commonly used for feature and edge extraction.

A Note on the Use of (Gray Level, Local Average Gray Level) Space as an Aid in Threshold Selection

A Note on the Use of (Gray Level, Local Average Gray Level) Space as an Aid in Threshold Selection