摘要：Complementary DNA (cDNA) microarrays are apowerful high throughput technology developed inthe last decade allowing researchers to analyze thebehaviour and interaction of thousands of genes si-multaneously. The large amount of information pro-vided by microarray images requires automatic tech-niques to develop accurate and efficient processing.Each spot in the microarray contains the hybridiza-tion level of a single gene. One of the most impor-tant features of these images are the regularity andpseudo-periodicity implicit in the spot arrangement.In this paper, an automatic approach based on tex-ture analysis characterization techniques is proposedto localize spots in microarray images. The methodestimates the displacement vectors which character-ize the texture (i.e. the spot arrangement). This isachieved by means of applying the generalized Houghtransform on the 2D autocorrelation function previ-ously segmented via morphological operations. Theobtained displacement vectors are used to generate agrid template which overlaps the original image. TheRoot-Mean-Square-Error (RMSE) between the esti-mated locations and the ones computed via a semiau-tomatic tool is calculated to evaluate the accuracy ofthe process. The method yields promising results
