This week was another productive week for wide-view scope development. There were several design changes (size, lumen diameter, and some feature tolerances) that I implemented, and resent the parts to be printed. The cameras also arrived from Hong Kong. These are tiny 2.8 mm in diameter endoscopic cameras with 140 \degrees and 120 \degrees FOV. The fit of the camera with the connector housing was good. I also wrote some scripts for an interface that allows me to visualize a real-time view of the camera, and also perform reconstruction on some of the frames, and view it on the side (Figure 1). The variable section is on the top right, which allows me to test different parameters that are needed for the reconstruction algorithm. One example is the duration of capture. For instance, if the duration of the capture time is 2 seconds, and once the acquisition is started, it considers all the continuous images acquired during the next 2 seconds, and uses them for reconstructions. I will be improving the image stitching method so that as the camera is slowly turned, it uses the images captured at different angles, into a 1 large image, which provides a "wider" view using the same FOV camera.
Figure 1: GUI for visualizing the camera real-time view (left window) and the reconstructed view (right window).
No comments:
Post a Comment