Confocal Microscopy

Confocal Microscopy: Images and Laser Scanning
The basic approach used to obtain an X-Y scan, is the combination of two oscillating mirrors mounted at right angles (see Scan Head mounting bracket). One mirror (M3H) oscillates in response to saw-tooth wave signal to form the vertical or X motion. The other mirror (CRS) oscillates or resonates as sine function to generate the horizontal scan or Y motion. Specialized Software: To obtain fast imaging rates both the forward and reverse scan of the CRS mirror are used. This requires that pixel information is collected during the reverse scan and is re-ordered before being saved as the final image. A inherent feature of a resonant scanner (CRS) is that the velocity of mirror movement slows to zero towards the end of each scan. A consequence of this motion in combination with a constant pixel clock (a feature common to image frame grabbers) is the spatial deformation of the image. However, this distortion is easily corrected by a sine function. This image distortion and the reversal of the pixel order of alternative line data can be corrected, in real-time, by a custom written application in Video Savant software as explained below.

(a) The original image is collected, in the frame grabber memory, at twice the final width and half the number of lines. Each line contains data from the forward and reverse scan and appears as a mirror image around the center (the end of the forward scan). (b) Because of the reduced velocity towards the ends of the scan, the image is distorted (center). The edges do not seem as as distorted because the scan is not recorded from 0 degrees. Image correction first requires that we change the real scan sequence shown in (c) to corrected image scan show in (d) by reversing and interlacing the data.
The correction of the image distortion is simply achieved using a sine function. The recorded linear position (squares) is divided by the radian value of the relative angular position of the scan (triangles) to give the corrected image position. This correction is performed in real-time.
Distorted image after reversal and interlacing of pixels (600 pixels x 400 lines)	Corrected Image (420 pixels x 400 lines)
Note: Not all 512 lines are used because time is required to return the vertical mirror and reset the frame grabber for each frame.