Illumination and Contrast Correction Strategy using Bilateral Filtering and Binarization Comparison

Illumination normalization and contrast variation on images are one of the most challenging tasks in the image processing field. Normally, the degrade contrast images are caused by pose, occlusion, illumination, and luminosity. In this paper, a new contrast and luminosity correction technique is developed based on bilateral filtering and superimpose techniques. Background pixels was used in order to estimate the normalized background using their local mean and standard deviation. An experiment has been conducted on few badly illuminated images and document images which involve illumination and contrast problem. The results were evaluated based on Signal Noise Ratio (SNR) and Misclassification Error (ME). The performance of the proposed method based on SNR and ME was very encouraging. The results also show that the proposed method is more effective in normalizing the illumination and contrast compared to other illumination techniques such as homomorphic filtering, high pass filter and double mean filtering (DMV).

[1]  Biao Wang,et al.  Illumination Normalization Based on Weber's Law With Application to Face Recognition , 2011, IEEE Signal Processing Letters.

[2]  Javier Ruiz-del-Solar,et al.  Illumination compensation and normalization in eigenspace-based face recognition: A comparative study of different pre-processing approaches , 2008, Pattern Recognit. Lett..

[3]  Nurulfajar Abd Manap,et al.  Disparity depth map layers representation for image view synthesis , 2014 .

[4]  Meng Joo Er,et al.  A Novel Local Illumination Normalization Approach for Face Recognition , 2011, ISNN.

[5]  Chris Dainty,et al.  Illumination correction of retinal images using Laplace interpolation. , 2012, Applied optics.

[6]  D. Ebenezer,et al.  Fast switching based median–mean filter for high density salt and pepper noise removal , 2014 .

[7]  Evangelos Dermatas,et al.  Non-uniform illumination correction in infrared images based on a modified fuzzy c-means algorithm , 2012 .

[8]  Hossein Shahamat,et al.  Face recognition under large illumination variations using homomorphic filtering in spatial domain , 2014, J. Vis. Commun. Image Represent..

[9]  James C. Gee,et al.  Retrospective illumination correction of retinal fundus images from gradient distribution sparsity , 2012, 2012 9th IEEE International Symposium on Biomedical Imaging (ISBI).

[10]  Ashish Kumar Bhandari,et al.  Improved knee transfer function and gamma correction based method for contrast and brightness enhancement of satellite image , 2015 .

[11]  Y. Cheng,et al.  Illumination Normalization Based on Different Smoothing Filters Quotient Image , 2010, 2010 Third International Conference on Intelligent Networks and Intelligent Systems.

[12]  Burairah Hussin,et al.  Image Duplication and Rotation Algorithms for Storage Utilization , 2015 .

[13]  Ajay Somkuwar,et al.  Image denoising and quality measurements by using filtering and wavelet based techniques , 2014 .

[14]  Oliver Speck,et al.  Prospective motion correction in brain imaging: A review , 2013, Magnetic resonance in medicine.

[15]  Qiuqi Ruan,et al.  An illumination normalization model for face recognition under varied lighting conditions , 2010, Pattern Recognit. Lett..

[16]  Alice Caplier,et al.  Illumination-robust face recognition using retina modeling , 2009, 2009 16th IEEE International Conference on Image Processing (ICIP).

[17]  Sazali Yaacob,et al.  Illumination normalization of non-uniform images based on double mean filtering , 2014, 2014 IEEE International Conference on Control System, Computing and Engineering (ICCSCE 2014).

[18]  Yu Cheng,et al.  Illumination normalization based on 2D Gaussian illumination model , 2010, 2010 3rd International Conference on Advanced Computer Theory and Engineering(ICACTE).

[19]  Liqiang Nie,et al.  Genetic algorithm and mathematical morphology based binarization method for strip steel defect image with non-uniform illumination , 2016, J. Vis. Commun. Image Represent..