This is a page of information about the 3-d snail location project, devised by Julian Derry and Chris Elliott. The snails (Lymnaea stagnalis, detailed view here) move around in a tank, viewed from the front and (using a mirror) the side. AVI-format files are generated using a Video camera and a Video-Blaster board. The analysis software (called (s)nail'em) is used to set simple criteria for characterizing snails and then each frame is scanned in turn using successively more exacting criteria. Eventually, the 3-d position is extracted from the data and recorded in a list box which can be output to a file. The software can be obtained here.
The example below shows screen shots of the software in action. Most of the screens have been reduced to save space and time! If these pictures are viewed with less than their normal palette of colors (e.g. on 16 color monitors) the snails may disappear!
First the AVI-format file is opened - the AVI file is shown in the left window: the right window is used for workspace.
The user chooses thresholds for the size of the snail and intensity of the light that discriminates between snail and background . In this he has the help of a pixel testing tool. Then a Dialog Box is used to set up for defining the tank outline
Next the tank (actual size) is outlined:
The user clicks on the Scan button and the frame is copied to the workspace window. The pixels are scanned and those which are as dark as the snail threshold are turned black; the others are white. In this example, the user has not quite got the tank selected right and part of the background is seen to the center of the Right hand window.
Finally the frame is examined again, pixel by pixel. White pixels are replaced by red as the scan progresses. The dark objects which are accepted as snails are colored green and added to the output list box. The rejected objects are colored yellow as shown in the insert (at actual screen size) where the edge of the tank is rejected.
The log file can be saved and analysed using an excel spreadsheet macro. The output from this macro is a 3 column table which can be imported into SPSS or Sigmaplot and viewed or further analysed. The results of one video are shown below
Send queries to firstname.lastname@example.org
November 14, 1995