Amount Of Blur Research Articles

Temporal integration of astigmatic blur information

Purpose: The accommodation response has recently been shown to improve visual acuity (VA) in the presence of astigmatism in some individuals. This improvement could potentially result from the accommodation system exhibiting a cyclic tracking response. This type of response would allow the integration of target features over time, resulting in an improvement in VA. In this study we measure VA in simulated astigmatic blur conditions, with the magnitude of astigmatic blur being temporally modulated over a range of astigmatic blur intervals. Improvement of VA with increased range of astigmatic blur would support the integration of visual information over time hypothesis.Methods: Three fully corrected subjects participated in a 3‐Alternative Force Choice method of constant stimuli paradigm where they were asked to identify one of three randomly chosen letters (K, R or X). The stimuli (Arial, font size 18) were subjected to varying amounts of Gaussian blur along the horizontal and vertical meridian and presented at a distance of 1.3 m. On any trial, the stimuli were randomly presented to the subjects at one of six astigmatic blur intervals. Each blur interval revolved at a temporal frequency of 2 Hz around the circle of least confusion with the smallest interval presenting the stimuli at the circle of least confusion and the largest oscillating the stimuli across the full interval, i.e. from the most horizontally to vertically blurred meridian.Results: From the resulting psychometric functions it was shown that all subjects exhibit a progressive improvement in letter recognition with increasing astigmatic blur interval.Conclusions: The improvement of VA with increased astigmatic blur interval strengthens the hypothesis that visual information can be integrated over a period of time.

Photographic reproduction of out-of-focus and distorted ocular imagery

The necessary settings and parameters were determined for ordinary camera and lens systems to faithfully reproduce out-of-focus and distorted imagery as it falls upon the retina of the human eye. Theoretic considerations of both geometric and physical optics were used to calculate the ‘relative blur’ and distortion produced by refractive error added to ordinary camera lenses as opposed to refractive error in an arbitrary thick-lens optical system bounded by air and fluid (i.e. the eye). In both the camera and the eye, ‘relative blur’ was determined to be directly proportional to dioptric defocus and to aperture size, and effectively independent of the focal length. Distortion of imagery was also found to be independent of the focal length. Photographs corroborate the theoretic findings. A given amount of relative blur, however, appeared somewhat greater when recorded on photographic film than when appreciated by the human eye. The Stiles–Crawford effect, the chromatic aberration of the eye, and neural processing probably each contribute to this difference. Previous investigators have grossly exaggerated blur and distortion in photographs intended to simulate ocular imagery and have drawn misleading conclusions from their results.

Objective Quality Potential Measures of Natural Color Images

This article deals with estimating the quality of natural color images from the digital image itself, for the purpose of determining whether the quality is sufficient for printing or displaying at some specified quality level. The problem of measuring simply defined parameters that correlate reasonably with human perception is discussed. Then an algorithm is described that measures sharpness, stochastic noise, and the blocking artifact of JPEG compression. The sharpness measure is the maximum of local sharpness, defined as the maximum gradient at an edge divided by the edge contrast. The algorithm behaves fairly consistently with respect to the amount of simulated blur and the measure correlates with the subjective sharpness of different images. The output of the stochastic noise algorithm is an unspecified number of noise levels, each of which is associated with the corresponding mean luminance. The algorithm computes a gradient image using the Sobel filters and uses the bivariate histogram of gradient and luminance to find the noise levels. The gradient filters are suitable for estimating the effect of power spectrum on noise visibility, but the discrimination between image details and noise is only partially solved by this method. The blocking artifact measure describes a characteristic spikiness in the histogram of the Sobel gradient.

No role for motion blur in either motion detection or motion-based image segmentation

The human contrast sensitivity function is bandpass in form for stimuli of low temporal frequency but low pass for flickering or moving stimuli. Because the loss in sensitivity to moving stimuli is large, images moving on the retina have little perceptible high-spatial-frequency content. The loss of high-spatial-frequency content--often referred to as motion blur--provides a potential cue to motion. The amount of motion blur is a function of stimulus velocity but is significant at velocities encountered by the visual system in everyday situations. Our experiments determined the influence of high-spatial-frequency losses induced by motion of this order on motion detection and on motion-based image segmentation. Motion detection and motion-based segmentation tasks were performed with either spectrally low-pass or spectrally broadband stimuli. Performance on these tasks was compared with a condition having no motion but in which form differences mimicked the perceptual loss of high spatial frequencies produced by motion. This allowed the relative salience of motion and motion-induced blur to be determined. Neither image segmentation nor motion detection was sensitive to the high-spatial-frequency content of the stimuli. Thus the change in perceptual form produced in moving stimuli is not normally used as a cue either for motion detection or for motion-based image segmentation in ordinary situations.

The effect of defocus on edge contrast sensitivity

The effect of optical blur (defocus) on edge contrast sensitivity was studied. Edge contrast sensitivity detoriates with fairly small amounts of blur (approximately 0.5 D) and is roughly reduced by half for each dioptre of blur. The effect of blur on edge contrast sensitivity equals the effect of blur on sine wave contrast sensitivity for a spatial frequency of approximately 3 cpd.

The effect of defocus on edge contrast sensitivity

The effect of optical blur (defocus) on edge contrast sensitivity was studied. Edge contrast sensitivity detoriates with fairly small amounts of blur (~0.5 D) and is roughly reduced by half for each dioptre of blur. The effect of blur on edge contrast sensitivity equals the effect of blur on sine wave contrast sensitivity for a spatial frequency of approximately 3 cpd.

Estimation of Edge Parameters and Image Blur Using Polynomial Transforms

A method is presented for detecting blurred edges in images and for estimating the following edge parameters: position, orientation, amplitude, mean value, and edge slope. The method is based on a local image decomposition technique called a polynomial transform. The information that is made explicit by the polynomial transform is well suited to detect image features, such as edges, and to estimate feature parameters. By using the relationship between the polynomial coefficients of a blurred feature and those of the a priori assumed (unblurred) feature in the scene, the parameters of the blurred feature can be estimated. The performance of the proposed edge parameter estimation method in the presence of image noise has been analyzed. An algorithm is presented for estimating the spread of a position-invariant Gaussian blurring kernel, using estimates at different edge locations over the image. First a single-scale algorithm is developed in which one polynomial transform is used. A critical parameter of the single-scale algorithm is the window size, which has to be chosen a priori. Since the reliability of the estimate for the spread of the blurring kernel depends on the ratio of this spread to the window size, it is difficult to choose a window of appropriate size a priori. The problem is overcome by a multiscale blur estimation algorithm where several polynomial transforms at different scales are applied, and the appropriate scale for analysis is chosen a posteriori. By applying the blur estimation algorithm to natural and synthetic images with different amounts of blur and noise, it is shown that the algorithm gives reliable estimates for the spread of the blurring kernel even at low signal-to-noise ratios.

Estimation of Edge Parameters and Image Blur Using Polynomial Transforms

A method is presented for detecting blurred edges in images and for estimating the following edge parameters: position, orientation, amplitude, mean value, and edge slope. The method is based on a local image decomposition technique called a polynomial transform. The information that is made explicit by the polynomial transform is well suited to detect image features, such as edges, and to estimate feature parameters. By using the relationship between the polynomial coefficients of a blurred feature and those of the a priori assumed (unblurred) feature in the scene, the parameters of the blurred feature can be estimated. The performance of the proposed edge parameter estimation method in the presence of image noise has been analyzed. An algorithm is presented for estimating the spread of a position-invariant Gaussian blurring kernel, using estimates at different edge locations over the image. First a single-scale algorithm is developed in which one polynomial transform is used. A critical parameter of the single-scale algorithm is the window size, which has to be chosen a priori . Since the reliability of the estimate for the spread of the blurring kernel depends on the ratio of this spread to the window size, it is difficult to choose a window of appropriate size a priori . The problem is overcome by a multiscale blur estimation algorithm where several polynomial transforms at different scales are applied, and the appropriate scale for analysis is chosen a posteriori . By applying the blur estimation algorithm to natural and synthetic images with different amounts of blur and noise, it is shown that the algorithm gives reliable estimates for the spread of the blurring kernel even at low signal-to-noise ratios.

Determination of surface roughness from reflected step edges

We suggest that a vision system may be able to determine surface roughness by measurement of the sharpness of the edges of an image reflected in a rough surface. We choose step edges because they are present in many environments and because the methods of edge detection and localization are mature. We show that the amount of blur of reflected images of step edges does indeed increase with respect to surface roughness. The method that we propose for estimating the roughness of a surface assumes a common surface-roughness model of simple form and requires a known shape for the reflective surface and a known position for the viewer. We show the results of experiments performed with planar milled surfaces in a controlled environment and calculate the roughness parameter. We also show that our method can rank order the surfaces by roughness. Next we compare our results with the values of the roughness as measured with a stylus profilometer to test the accuracy of our method and of the model. We find that the results of the two methods are quite different. We discuss the possibility that this difference is due to the non-Gaussian character of the surface height distribution and describe current research into possible models of the surfaces. We conclude by issuing a warning to computer vision researchers that the most common model of surface roughness might be inadequate for describing many real surfaces.

Stability of Oculomotor Fixation as a Function of Target Contrast and Blur

Previous studies have shown that changes in target parameters, such as luminance, color, size, and shape exert little effect on the stability of fixation, measured as the standard deviation of eye positions over several seconds of fixation. In this study, we investigated the effect of target contrast and dioptric blur on the stability of fixation in five normal subjects. The horizontal positions of both eyes were assessed by infrared limbal tracking while the subjects fixated binocularly for 20-s intervals on a cross-shaped target at 2 m. Across subjects, the mean stability of fixation on a high contrast target worsened significantly from 8.9 to 12.0 min arc with fixation through increasing amounts (0 to 4 D) of dioptric blur. However, without blur there was no statistically significant effect on fixation stability of varying the target contrast from 7 to 84%. The variability of vergence was not significantly influenced in either condition. Like other target parameters investigated in previous studies, dioptric blur and contrast have relatively inconsequential effects on fixation stability, indicating that the signals used to control fixation are largely independent of stimulus parameters. Practically, the findings suggest that when patients' eye movements are recorded for clinical reasons, the results are unlikely to be affected seriously by the presence of residual refractive error.

Depth perception using blurring and its application in VLSI wafer probing

In this paper, we present a technique for measuring the amount of blur of an edge and using this information to determine the distance of a micromanipulator probe from a wafer surface in very large scale integration (VLSI) wafer probing. In this application, a soft and reliable touch of the probe with a metal pad in the wafer is a sensitive operation. The wafer is focused and several images of the probe while approaching the wafer are analyzed. In our theory, the amount of blur is calculated from the height of the step edge and the slope of the intensity profile at the zero crossing. Hence, our formula is simple and easy to use. We estimate the distance of the probe from the surface of the wafer and obtain a robust measure, i.e., one which is valid even in the presence of significant noise in the images. In order to validate our methods, we have experimented with various VLSI patterns as backgrounds.

On the Feasibility of Cross-Validation in Image Analysis

The performance of cross-validation as a device for selecting the optimal amount of regularisation in image analysis is examined. It is shown that cross-validation results in asymptotically optimal performance, provided the amount of blur does not exceed a certain ceiling. If blur exceeds that level, then cross-validation does not yield asymptotic optimality. The ceiling is surprisingly low. For example, if a fixed scene is subject to blur and is digitised on an increasingly fine grid, then cross-validation will only yield optimal image enhancement if the amount of blur decreases as the grid becomes finer. Two types of cross-validation are studied, one taken directly from methodology for nonparametric regression, and the other involving a modification that adjusts for image blur.

3-D position sensing using a passive monocular vision system

Passive monocular 3-D position sensing is made possible by a new calibration scheme that relates depth to focus blur through a composite lens and aperture model. The calibration technique enables the recovery of absolute 3-D position coordinates from image coordinates and measured focus blur. A geometric model of the camera's position and orientation in space is used to transform the camera's imaging coordinates into world coordinates. The relationship between the world coordinate system and the screen coordinate system which includes the amount of focus blur, is developed by modeling the camera imaging arrangement. The modeling proceeds first through the perspective view from a pinhole camera located anywhere in space. The camera's lens and aperture system is investigated to find the relationship between depth and focus blur. The aspect ratio of the frame image is considered. Position accuracies comparable to those in stereo based vision systems are possible without the need for solving the difficult point of correspondence problem.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>

Color dependent optical prefilter for the suppression of aliasing artifacts

Optical prefilters, which make use of the double-refraction effect in crystalline quartz, reduce the amount of aliasing artifacts in the displayed image of a video system. This paper describes anew class of optical prefilter that produces different amounts of blur for different colors. The chromatic variation is obtained by placing a polarization retarder between two quartz crystals. This type of filter is useful with sensors that have unequal pixel densities in the three sensed colors.

Speechreading with reduced vision: a problem of aging.

Presbycusis, a progressive loss of hearing in the elderly, causes people to become dependent on vision for understanding spoken language at the same time that visual disorders of aging compromise their vision. This combination of sensory losses disrupts the lives of millions of elderly Americans but remains virtually unstudied. We tested the ability of hearing and hearing-impaired adults to understand sentences spoken by an actress without sound. She was seen through various amounts of dioptric blur. Both young and elderly adults were relatively unaffected by as much as 4 D of blur. They quickly adjusted to new levels of blur, showing improved performance within a few minutes. Ocular pathologies altered the effect of blur on speechreading.

Television and vision: reading captions when vision is blurred.

Closed captions for television have become an important means of communication for hearing-impaired people. Yet, previous studies show that hearing-impaired people are very likely to have visual problems that may interfere with caption reading. In this study, the effect of reduced vision on caption reading is examined. Six hearing-impaired and 14 hearing subjects viewed segments of captioned movies on a 19-inch television screen at a distance of three meters. All of the subjects had normal vision. Their vision was blurred with plus lenses of five different strengths. The data show that caption reading is affected by relatively small amounts of blur. If a person has even a slight visual problem, the caption letters under normal television viewing conditions are too small for maximum reading performance.

Measurement Of The Landsat Thematic Mapper Modulation Transfer Function Using An Array Of Point Sources

This paper presents a method for measuring the Thematic Mapper (TM) imaging system point spread function (PSF) using TM imagery or a specially constructed target consisting of a two-dimensional array of approximate point sources of known dimensions and radiometric qualities. The target allows 16 separate point sources to be imaged simultaneously by the TM. The point sources were carefully placed on the ground so that their relative positions were known. Owing to sample-scene phasing, each imaged point source exhibits a different amount of blur in the digital image. The target pixels may then be recombined according to their known relative positions to form a single, sampled, nonaliased imaging system PSF. The modulation transfer function is then obtained as the modulus of the discrete Fourier transform of the PSF.

Reconstruction of blood vessels from x-ray subtraction projections: limited angle geometry.

Several algorithms have been investigated for reconstructing blood vessels from a limited number of x-ray subtraction projections, distributed over a limited range of angles. Both computer simulations and an in vivo animal study were carried out. The best reconstruction performance was achieved using an algorithm that folded in two pieces of a priori knowledge of the vascular density distributions: (1) the object is dilute, consisting mainly of a void; and (2) the density distribution in the reconstructions is most likely to be non-negative. Both the signal-to-noise ratio (SNR) and the signal to out-of-focus blur were quantitated. Compared to tomosynthetic reconstruction (backprojection), the amount of residual blur from out-of-focus planes was significantly reduced with only a small penalty in diminished SNR. The combined effect resulted in significant qualitative image improvement for real arterial distributions as demonstrated in a canine arterial imaging example.

Performance of the Dot Visual Acuity Test with Uncorrected Ametropia Resulting in Target Blur

The dot visual acuity test is used to test the acuity of children between the ages of 2 and 4 years who do not respond to the tumbling E test. Because the test involves the detection of a black dot on a white background it is appropriate to ask if the detection threshold is lowered artificially in the presence of target blur. Twenty-one children (42 eyes) were used in the study. Refractive errors were fully corrected and each subject's lag of accommodation was measured. A monotonic decrease in acuity was found with increasing amounts of optical blur. The explanation for the decreased acuity is found in the reduced retinal image contrast of the blurred target.

Image Sharpness Enhancement Using Adaptive 3X3 Convolution Masks

Biased Laplacian masks are useful for enhancing the sharpness of edges in images degraded by moderate amounts of blur and noise. Further-more, their implementation on pipelined processors is fast and convenient. However, as we show, they become limited by noise very rapidly. The use of a mask whose bias adapts to local variance measured in 3 X3 windows is explored in this paper. Results demonstrate a significant enhancement of the power of biased Laplacian masks.