Automatic Gridding of Rotated Microarray Images

This paper presents a new method for automatically gridding rotated microarray images. In previous methods rotation of an image is either ignored or a high amount of user intervention is required to correct the problem. Automation of this process is essential. The proposed method uses the Hough Transform to find the current orientation of the image. Once the orientation is found the microarray can be automatically rotated to allow gridding to be performed. The proposed method works independent of spot colour, size and location, thus the number of user parameters is kept at a minimum. Most notably the program performed rotations on microarrays with a rotation of 0.6 degrees. The proposed method can allow processing of images scanned in any orientation and as a result greater throughput can be achieved.