سال انتشار: ۱۳۹۰

محل انتشار: هفتمین کنفرانس ماشین بینایی و پردازش تصویر

تعداد صفحات: ۵

نویسنده(ها):

Seyed Pooya Ehsani – Depatment of Electrical Engineering, Sharif University of Technology, Tehran, Iran
Hojjat Seyed Mousavi – Depatment of Electrical Engineering, Sharif University of Technology, Tehran, Iran
Babak.H. Khalaj – Depatment of Electrical Engineering, Sharif University of Technology, Tehran, Iran

چکیده:

Vivid banding patterns in medical images ofchromosomes are a vital feature for karyotyping and chromosomeclassification. The chromosome image quality may be degradedby many phenomenon such as staining, sample defectness andimaging conditions. Thus, an image enhancement processingalgorithm is needed before classification of chromosomes. In thispaper, we propose an adaptive and iterative histogram matching(AIHM) algorithm for chromosome contrast enhancementespecially in banding patterns. The reference histogram, withwhich the initial image needs to be matched, is created basedon some processes on the initial image histogram. Usage ofraw information in the histogram of initial image will resultin more dependency to the input image and acquiring bettercontrast improvement. Moreover, the iteration procedure leads toa gradual contrast enhancement and getting the best result. Theiteration steps may vary depending on the image characteristicsand histogram. In order to assess the performance of the proposedalgorithm in comparison with existing image enhancementtechniques, Constant Gain Transform (CGT) and Local StandardDeviation Adaptive Contrast Enhancement (LSD-ACE), aquantitative measurement, the contrast improvement ratio (CIR),is utilized. The experimental results indicate that the proposedmethod shows the best results in terms of the CIR measure and,as well as in visual perception.