(For USM Staff/Student Only)

EngLib USM > Ω School of Electrical & Electronic Engineering >

Three dimension pixel distribution shifting based color correction method for digital images

Three dimension pixel distribution shifting based color correction method for digital images / Goh Jinn Chy
Algorithma warna pembetulan untuk imej digital warna adalah penting untuk mengurangkan pengaruh pencahayaan sekeliling yang tidak sekata dalam imej digital warna. Biasanya, algoritma-algoritma warna pembetulan dibahagikan kepada dua fasa utama iaitu penganggaran warna dan proses pembetulan warna. Walaupun terdapat pelbagai algoritma warna pembetulan telah diperkenalkan oleh penyelidik-penyelidik lain tetapi imej yang dihasilkan masih menghadapi beberapa masalah. Beberapa algoritma masih menghasilkan imej berwarna yang tidak sekata, manakala kaedah-kaedah lain masih menghasilkan imej-imej yang terlebih pembetulan. Oleh itu, dalam kajian ini, algoritma warna pembetulan baharu akan diperkenalkan untuk mengatasi kelemahan-kelemahan algoritma konvensional. Algoritma diperkenalkan dalam kajian ini dinamakan sebagai Tiga Dimensi Pembetulan Warna Anjakan Piksel (3D-PWAP). Di dalam kajian ini, dua jenis 3D-PWAP akan diperkenalkan iaitu pertama 3D-PWAP berdasarkan saluran warna RGB dan 3D-PWAP kedua berdasarkan saluran warna CIE*Lab. Kedua-dua algoritma yang dicadangkan menggunakan proses pembetulan warna dengan peralihan pengedaran piksel berdasarkan pada satah 3D tiga saluran warna. Daripada keputusan analisis kualitatif, algoritma 3D-PWAP berdasarkan saluran warna RGB dapat menghasilkan imej yang lebih baik bagi imej di bawah laut. Kebiruan imej dikurangkan dengan ketara dan warna objek yang ditangkap dapat ditentukan dengan lebih mudah. Algoritma 3D-PWAP berdasarkan saluran warna CIE*Lab adalah lebih sesuai untuk membetulkan imej yang mempunyai kawasan terlebih pendedahan. Algoritma ini memastikan kehilangan maklumat tidak akan berlaku semasa proses pembetulan imej. Selain itu, analisis kuantitatif juga menunjukkan bahawa algoritma yang dicadangkan bukan sahaja mempunyai keupayaan untuk membetulkan warna imej, tetapi juga dapat mengekalkan kecerahan imej. _______________________________________________________________________________________________________ Color correction algorithm for a color digital image is important to reduce the influence of inhomogeneous surrounding illumination in color digital image. Commonly, existing color correction algorithms are divided into two main phases, namely color estimation and color correction. Although there are various color correction algorithms introduced by other researchers, the output images created still face some problems. Some algorithms still produce non-uniform illumination resultant image, while others methods still produce over-corrected problem in the resultant images. Thus, in this study, new color correction algorithm was developed to overcome the weaknesses of those conventional algorithms. The introduced algorithm is named as Three Dimension Pixel Shifting Color Correction (3D-PSCC). There are two variations of 3D-PSCC, which are 3D-PSCC based on RGB color channel and 3D-PSCC based on CIE*Lab color channel. Both proposed algorithms employ color correction process by shifting the pixel distribution based on the 3D three color channels planes. From the qualitative analysis results, the 3D-PSCC algorithm based on RGB color channels is able to produce better resultant images for underwater images. The bluish of the images are reduced significantly and the color of the object captured is able to be determined easily. The 3D-PSCC algorithm based on CIE*Lab color channel, on the other hand, is more suitable to correct the overexposed images. This algorithm ensures loss of information does not occur during image enhancement process. Besides that, the quantitative analysis also proves that the proposed algorithms introduced not only have the ability to correct the color of the image, but are also able to remain the brightness of the image.
Contributor(s):
Goh Jinn Chy - Author
Primary Item Type:
Final Year Project
Identifiers:
Barcode : 00003107032
Accession Number : 875007154
Language:
English
Subject Keywords:
Color correction algorithm; inhomogeneous surrounding illumination; color digital image
First presented to the public:
6/1/2017
Original Publication Date:
4/18/2018
Previously Published By:
Universiti Sains Malaysia
Place Of Publication:
School of Electrical & Electronic Engineering
Citation:
Extents:
Number of Pages - 88
License Grantor / Date Granted:
  / ( View License )
Date Deposited
2018-04-18 17:22:06.6
Date Last Updated
2019-01-07 11:24:32.9118
Submitter:
Mohd Jasnizam Mohd Salleh

All Versions

Thumbnail Name Version Created Date
Three dimension pixel distribution shifting based color correction method for digital images1 2018-04-18 17:22:06.6