Effects Of Non-uniform Illumination Research Articles

블록 기반 밝기 표준화를 통한 이진영상의 고속 불균일 조명 보정

We investigated a novel fast non-uniform illumination correction method for bi-level images. The proposed method divides a bi-level image into sub-images and roughly estimates block-wise illumination by low pass filtered maximum values of sub-images. After that, we apply bilinear interpolation using the block-wise illumination to estimate non-uniform illumination, and compensate for the effect of non-uniform illumination using the estimated illumination. Since the proposed method is not based on computation intensive iterative optimization, the proposed method can be used effectively for applications that require fast correction of non-uniform illumination. In simulations, the proposed method showed more than 20 times faster speed than existing entropy minimization method. Moreover, in simulations and experiments, the restored images by the proposed method were more close to true images than images restored by conventional method.

Non-uniform illumination correction in infrared images based on a modified fuzzy c-means algorithm

The correction of non-uniform illumination and the elimination of shading artifacts is an important preprocessing task used in a great number of image processing applications such as segmentation, registration or quantitative analysis. Although, a careful and accurate set up of the image acquisition system may degrade the importance of a brightness normalization algorithm, non-uniform illumination appears due to the interaction of objects and light on the scene requires retrospective shading correction. The image formation process and the corresponding shading effects are described by a linear image formation model, which consists of a multiplicative and an additive shading component. In this paper a novel brightness normalization method is proposed to eliminate the non-uniform illumination effects. The method is based on the application of a fuzzy c-means algorithm (FCM) only on the background part of the acquired image, where the objective function is modified to take into account local information of each pixel in the estimation of the multiplicative and the additive shading components. The modified FCM algorithm is iterative, as the standard FCM, and at each iteration the multiplicative and the additive shading components are re-estimated based on the cluster centers and the membership of each pixel in a specific cluster. Brightness correction is performed by the inverse of the image formation model after FCM convergence. Experiments were conducted in a database of both real and artificial infrared images. The experimental results show that the proposed method decreases significantly the non-uniform illumination effects and does not introduce brightness variations if the background is uniform.

Comparative study of retrospective methods to reduce non-uniform illumination effects to bridge coating

Abstract Digital image processing methods have been actively studied and applied to various infrastructure components including bridge painting assessment. Despite the advantages of image recognition techniques to bridge coating, the results can be greatly affected by the non-uniform illumination of target conditions. It is necessary to first convert all images to a standardized background illumination condition. The effect caused from non-uniform illumination can be minimized by setting lighting conditions properly or capturing additional calibration images. However, capturing additional calibration images is not possible in many practical situations and hence a good method of shading correction is required. This paper addresses the way to correct digital bridge coating images containing non-uniform illumination, which becomes more difficult when it comes to assessing facility surfaces with rust defects. In order to choose the most appropriate one, four representative shading correction methods were selected and tested to identify their suitability to processing bridge coating images. These methods include: (1) image smoothing, (2) homomorphic filtering, (3) morphological filtering, and (4) entropy minimization method. Test results on bridge coating images showed that the entropy minimization method outperformed the other three methods.

Fourier Transform based texture measures for grinding wheel condition monitoring using machine vision

The surface quality of the component produced mainly depends on the condition of the grinding wheel. In this paper, the working surface condition of the grinding wheel is assessed by the texture analysis methods. The grinding wheel images are captured at different time intervals using a CCD camera. Then, the captured images are pre-processed using histogram equalisation to reduce the effect of non-uniform illumination. After pre-processing, Fourier power spectrum based texture parameters are evaluated and used for discriminating the grinding wheel condition. The evaluated parameters identify the condition of the grinding wheel efficiently, which in turn provide an opportunity for online condition monitoring of grinding wheel using machine vision.

Robust iris indexing scheme using geometric hashing of SIFT keypoints

This paper proposes an efficient indexing scheme for searching large iris biometric database that achieves invariance to similarity transformations, illumination and occlusion. The proposed scheme considers local descriptors as well as relative spatial configuration for claiming identity. To overcome the effect of non-uniform illumination and partial occlusion due to eyelids, local features are extracted from noise independent annular iris image using scale invariant feature transform (SIFT). The detected keypoints are used to index iris database by applying geometric hashing scheme that is robust to similarity transformations as well as occlusion. During iris retrieval, geometric hashed location from query iris image is obtained to access the appropriate bin of hash table and for every entry found there, a vote is casted. The iris images that receive more than certain number of votes are considered as possible candidates. In order to find the potential matches, the keypoint descriptor of the list of possible candidates is matched with the query iris. Since only small portion of database is scanned to find a match it reduces the query retrieval time and improves accuracy. This approach is tested on UBIRIS, BATH, CASIA and IITK iris databases and shows a substantial improvement over exhaustive search technique in terms of time and accuracy.

Retrospective correction of nonuniform illumination on bi-level images

We propose a novel method for correcting the effect of nonuniform illumination on a bi-level image. The proposed method is based on a penalized nonlinear least squares objective function that measures the binariness of an image and the roughness of illumination. Compared with conventional methods, it has the advantages of 1) not suffering from a trivial minimizer, 2) not requiring tuning of design parameters, and 3) effective optimization. In addition, it yields a unique solution since the minimization of the objective function is well-posed. In simulations and experiments, the method showed better accuracy and speed than the conventional entropy-based method.

A two-dimensional finite element model of front surface current flow in cells under non-uniform, concentrated illumination

A two-dimensional finite element model of current flow in the front surface of a PV cell is presented. In order to validate this model we perform an experimental test. Later, particular attention is paid to the effects of non-uniform illumination in the finger direction which is typical in a linear concentrator system. Fill factor, open circuit voltage and efficiency are shown to decrease with increasing degree of non-uniform illumination. It is shown that these detrimental effects can be mitigated significantly by reoptimization of the number of front surface metallization fingers to suit the degree of non-uniformity. The behavior of current flow in the front surface of a cell operating at open circuit voltage under non-uniform illumination is discussed in detail.

Deformation image generation for testing a strain measurement algorithm

An optical extensometer was tested using artificially deformed images with a known strain field. A real image series from a tensile test was used to obtain realistic deformation parameters, including spatial and temporal strain characteristics, changes in tonal pixel properties due to deformation, and the effect of nonuniform illumination. These parameters are used to artificially deform a real image taken from an object with a random speckle pattern. The signal-to-noise ratio of the resulting artificially deformed images is varied by applying a blurring pillbox filter and additive Gaussian noise to them. The optical extensometer uses digital image correlation to track homologous points of the object, and further to measure strains. The strain measurement algorithm includes a heuristic to dynamically control the template size in image correlation. Furthermore, several other methods to improve the accuracy-complexity ratio of the algorithm exist. The effects of different parameters and heuristics on the accuracy of the algorithm as well as its robustness against blur and noise are studied. Results show that the proposed test method is practical, and the heuristics improve the accuracy and robustness of the algorithm.

Embedded bone fragment detection in chicken fillets using transmittance image enhancement and hyperspectral reflectance imaging

This paper is concerned with the detection of bone fragments embedded in compressed de-boned skinless chicken breast fillets by enhancing single-band transmittance images generated by back-lighting and exploiting spectral information from hyperspectral reflectance images. Optical imaging of chicken fillets is often dominated by multiple scattering properties of the fillets. Thus, resulting images from multiple scattering are diffused, scattered and low contrast. In this study, a fusion of hyperspectral transmittance and reflectance imaging, which is a non-ionized and non-destructive imaging modality, was investigated as an alternative method to the conventional transmittance X-ray imaging technique which is an ionizing imaging modality. An image formation model, called an illumination–transmittance model, was applied for correcting non-uniform illumination effects so that embedded bones are more easily detectable by a simple segmentation method using a single threshold value. Predicted bones from the segmentation were classified by the nearest neighbor classifier that was trained by the spectral library of mean reflectance of chicken tissues like fat, meat and embedded bones. Experimental results with chicken breast fillets and bone fragments are provided.

Facial feature extraction using complex dual-tree wavelet transform

In this paper, we propose a novel method for facial feature extraction using the directional multiresolution decomposition offered by the complex wavelet transform. The dual-tree implementation of complex wavelet transform offered by Selesnick is used (DT-DWT(S)) [I.W., Selesnick, R.G. Baraniuk, N.C. Kingsbury, The dual-tree complex wavelet transform, IEEE Signal Processing Magazine, 6, s.l., IEEE, November 2005, vol. 22, pp. 123–151.]. In the dual-tree implementation, two parallel discrete wavelet transform (DWT) with different lowpass and highpass filters in different scales are used. The linear combination of subbands generated by two parallel DWT is used to generate 6 different directional subbands with complex coefficients. A test statistic, which is derived with absolute value of complex coefficient, whose distribution matches very closely with the directional information in the 6 subbands of the DT-DWT(S) is derived and used for detecting facial feature edges. The use of the complex wavelet transform is motivated by the fact that it helps eliminate the effects of non-uniform illumination, and the directional information provided by the different subbands makes it possible to detect edge features with different directionalities in the corresponding image. Edge information of facial area is enhanced using multiresolution structure of DT-DWT(S). The proposed method also employs an adaptive skin colour model instead of a predefined skin colour statistic. The model is developed with a unimodal Gaussian distribution using the skin region which is extracted excluding the detected edge map obtained from the DT-DWT(S). By combining the edge information obtained by using DT-DWT(S) and the non-skin areas obtained from the pixel statistics, the facial features are extracted. The algorithm is tested over the well known Carnegie Mellon University (CMU) and Marks Weber face databases. The average detection rate of the proposed method using DT-DWT(S) provides up to 9.6% improvement over the same method using discrete wavelet transform (DWT).

Bone Fragment Detection in Chicken Breast Fillets Using Transmittance Image Enhancement

This article is concerned with the detection of bone fragments embedded in de-boned skinless chicken breast fillets by modeling optical images generated by backlighting. Imaging of chicken fillets is often dominated by multiple scattering properties of the fillets. Thus, resulting images from multiple scattering are diffused, scattered, and low contrast. In this study, a combination of transmittance and reflectance hyperspectral imaging, which is a non-ionized and non-destructive imaging modality, was investigated as an alternative method to conventional transmittance x-ray imaging, which is an ionizing imaging modality. As a way of reducing the influence of light scattering on images and thus increasing the image contrast, the use of a structured line light was examined along with an image formation model that separated undesirable lighting effects from an image. The image formation model, based on an illumination-transmittance model, was applied for correcting non-uniform illumination effects so that embedded bones were more easily detected by a single threshold. An automated image processing algorithm to detect bones was also proposed. Experimental results with chicken breast fillets and bone fragments are provided. The detection accuracy of the developed technology was 100%. The false-positive rate was 10%.

Effectiveness of neuro-fuzzy recognition approach in evaluating steel bridge paint conditions

The development of digital image recognition techniques has contributed to increased precision in pattern recognition and led to numerous applications in industries. In September 1999, the Indiana Department of Transportation (INDOT) first tried out digital image recognition techniques to steel bridge coating assessment. The purpose of this tryout was to obtain a rust percentage, which was required in the INDOT bridge painting warranty contract, when conducting steel bridge coating investigation. Despite the advantages of digital image recognition, some problems that may cause inaccurate recognition results still exist. Nonuniform illumination (i.e., brightness or darkness or shadow) is one of them. The neuro-fuzzy recognition approach (NFRA) was developed to minimize the effect of nonuniform illumination. In this technical note, the framework of NFRA, its application to steel bridge coating assessment, and its performance comparison to three other image recognition methods will be presented.Key words: neuro-fuzzy recognition approach (NFRA), artificial neural network (ANN), double sampling plan, multiresolution pattern classification (MPC), iterated conditional modes (ICM).

Two-dimensional simulations of plastic-shell, direct-drive implosions on OMEGA

Multidimensional hydrodynamic properties of high-adiabat direct-drive plastic-shell implosions on the OMEGA laser system [T. R. Boehly et al., Opt. Commun. 133, 495 (1997)] are investigated using the multidimensional hydrodynamic code, DRACO [D. Keller et al., Bull. Am. Phys. Soc. 44, 37 (1999)]. Multimode simulations including the effects of nonuniform illumination and target roughness indicate that shell stability during the acceleration phase plays a critical role in determining target performance. For thick shells that remain integral during the acceleration phase, target yields are significantly reduced by the combination of the long-wavelength (ℓ<10) modes due to surface roughness and beam imbalance and the intermediate modes (20⩽ℓ⩽50) due to single-beam nonuniformities. The neutron-production rate for these thick shells truncates relative to one-dimensional (1D) predictions. The yield degradation in the thin shells is mainly due to shell breakup at short wavelengths (λ∼Δ, where Δ is the in-flight shell thickness). The neutron-rate curves for the thinner shells have significantly lower amplitudes and a fall-off that is less steep than 1D rates. DRACO simulation results are consistent with experimental observations.

Limit of the Surface-Color Mode Perception under Non-Uniform Illuminations

A series of experiments were conducted to find the effects of non-uniform illumination on the surface-color mode perception. Two patterns of the illumination, one-sided illumination and a spotlight, were simulated. Observers adjusted the luminance of the test stimulus so that it just started to appear partially as an aperture-color mode. We found that the upper-limit luminance was significantly lower for all test colors when the directions of the gradient between the test stimulus and the surrounds did not match. On the other hand, in the spotlight conditions the upper-limit luminances changed only when it was contained in the spotlighted area. Our results suggest that the brightest stimulus in the scene does not work as a cue, and that the visual system takes the influence of illumination into account in order to set a criterion for the judgment for the color appearance of the mode.

Effects of Nonuniform Illumination on Implosion Asymmetry in Direct-Drive Inertial Confinement Fusion

Target areal density (rhoR) asymmetries in OMEGA direct-drive spherical implosions are studied. The rms variation <deltarhoR>/<rhoR> for low-mode-number structure is approximately proportional to the rms variation of on-target laser intensity <deltaI>/<I> with an amplification factor of approximately 1/2(C(r)-1), where C(r) is the capsule convergence ratio. This result has critical implications for future work on the National Ignition Facility as well as OMEGA.

Determining the hydrographic parameters of the surface of water from the image sequences of a CCD camera

Water surface fluctuations were analyzed using image sequences from a CCD (charge coupled device) camera. Both regular and irregular waves were used for the study. A transfer function is obtained by comparing the gray scales of the image and results of “in-situ” wave gauge array measurements. The effects of non-uniform illumination within the wave basin are imbedded in the transfer function. Thus, the obtained transfer function is directly applicable to all other areas of interests within the wave basin. Estimated wave heights are all in qualitative agreement with those obtained through direct measurements. It is shown that the averaged relative error between these two wave heights is less than 16%.

Characterization of port wine stain skin erythema and melanin content using cross-polarized diffuse reflectance imaging.

Objective methods to assess quantitatively port wine stain (PWS) blanching in response to laser therapy are needed to improve laser therapeutic outcome. Previous studies have attempted to assess objectively PWS color based on point measurement devices. To date, these approaches have typically been limited by a number of factors such as small test area and need for contact. To address these issues, a cross-polarized diffuse reflectance imaging system and color image analysis method has been developed to evaluate quantitatively erythema and melanin content in PWS skin. A cross-polarized diffuse reflectance system has been constructed to acquire high resolution digital images while minimizing artifacts such as glare, shadowing, and nonuniform illumination effects that can compromise image fidelity. Furthermore, an image analysis algorithm has also been developed to analyze normal and PWS skin in terms of CIEL*a*b* color space parameters. Using the algorithm, images of color space intensity, L*, and saturation, a*, indices have been calculated to extract quantitative metrics of melanin and erythema, respectively. Compared to the cross-polarized diffuse reflectance image, the a* index image enhanced the contrast in regions of high and low erythema and seems to have quasi correlation with the L* index image. In the presented PWS patient example, the relative a* index difference (Deltaa*) image showed less erythema in PWS skin after laser treatment when compared to pre-treatment values. Our cross-polarized imaging system and color image analysis method is a simple, noncontact technique that can indirectly provide quantitative measurement of erythema and melanin content in PWS skin.

Removing non-uniform illumination effects from deep-sea floor images

The deep-sea floor images are acquired under non-uniform illumination, the effects of which bring up many difficulties for image segmentation. The paper compares three methods of removing non-uniform illumination effects. The effectiveness and robustness are evaluated on three test images with high-, moderate- and low-density grains, respectively. The results show that local filtering algorithm does not remove the non-uniform illumination effects completely. The image subtracting algorithm will lose some of the dynamic range of the original data. The enhanced image stretching algorithm is the most effective one.

Image registration for pressure-sensitive paint applications

Pressure-sensitive paint (PSP) data are generally derived by ratioing two images of a PSP-coated wind tunnel model. One image is taken at a known “wind-off” reference pressure while the other is taken at the test condition, in order to factor out the effects of nonuniform illumination and paint thickness. However, airloads on the model cause it to move during testing with respect to both the camera and the illumination source used to excite the P & P. Thus, the raw PSP images must be aligned, or registered, by means of a spatial transform. Once a proper ratioed image is obtained, image coordinates must be related to model coordinates so that pressure data can be obtained at the desired points on the model. This article compares several transforms used for image registration, and demonstrates the use of the projective equation of photogrammetry to relate model to image coordinates. An application in a small-scale wind tunnel test is shown, demonstrating mapping of pressure data to the three-dimensional model surface.

Power lines: radar measurements and detection algorithm for polarimetric SAR images

Polarimetric radar backscattering measurements of a variety of powerline cables are presented. The objective of the first part of the investigation was to study the effect of braiding of the cables on the backscattering at skew incidence. The measurements were performed for four different actual size powerline samples at C-, X-, and Ka-band over a wide range of incidence angles. The data were collected over a 500 MHz bandwidth at C- and X-band with a 1.25 MHz increment and a 1 GHz bandwidth at Ka-band with a 2.5 MHz increment. Also the effect of nonuniform illumination and measurement in the near field of the cables were studied. Experimental data shows a significant radar backscatter for W-polarization (/spl sigma//sub /spl nu//spl nu//) at angles away from normal incidence. This backscatter is proportional to the number and diameter of the strands on the surface of the cables. There is also noticeable backscatter for the HH and VH components of the scattering matrices. Their magnitudes, relative to that of the W component, are proportional to the pitch angle of the helix. Since detection of these cables is an important safety issue for low-flying airplanes a detection algorithm using polarimetric synthetic aperture radar (SAR) images was developed using the knowledge gleaned from the measurements. The detection algorithm was tested on a simulated image and worked well, detecting a power line whose backscatter power was 6 dB below the average background power.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>