Automated chromosome counting with color and grayscale microscope images
Abstract
We present a system that can automatically count the number of chromosomes present in a metaphase spread. Image processing techniques on huesaturation-value (HSV) and normalized RG (rgI) color spaces and grayscale images are utilized in the detection of each chromosome. Upon comparing the performance of each technique, it is shown that operating in grayscale gives the best results. Testing the system with nine chromosome metaphase spread samples, an average success rate of 94% for grayscale images, 89% for HSV, and 93% for rgI is obtained.