Anne & Zener GA
Very Active Member
Yesterday
Woohoo! :mrgreen: :mrgreen: :mrgreen: I tested again at PS and got a 61. Appy is good. Zener did a little yard patrol and is now napping on the filing cabinet helping with his condo. I shot the s1.0u, which is the same dose (reduction taken yesterday morning from 1.0u). We can be around to monitor today.
Liz
OT: There is a group of electrical engineering students at my university working on a dosing app for the iPad. Here is a video of their progress. The explanation is below.
http://youtu.be/-bfVqLz5XX0
The first part of the video shows an example of the many individual photos needed to develop classifier files for each feature to be detected.
These files are used to locate objects from within an image.
Next you can see the first iteration of the syringe detection software.
The entire frame is being processed leading to the slow frame rate.
False positives are also visible in the far upper left corner.
The improved feature detection limits the processed region to a small box in the center of the visible field.
This method reduces false positives while improving the processing time.
This improved processing is shown in the improved fame rate of the video.
The last demonstration is of a simple feature displacement algorithm.
The two features coordinates are compared to one another and the hypotenuse between the points is displayed and measured in pixels.
The next step would be to allow for accurate scaling of the image to compensate for camera distance and parallax.
Woohoo! :mrgreen: :mrgreen: :mrgreen: I tested again at PS and got a 61. Appy is good. Zener did a little yard patrol and is now napping on the filing cabinet helping with his condo. I shot the s1.0u, which is the same dose (reduction taken yesterday morning from 1.0u). We can be around to monitor today.
Liz
OT: There is a group of electrical engineering students at my university working on a dosing app for the iPad. Here is a video of their progress. The explanation is below.
http://youtu.be/-bfVqLz5XX0
The first part of the video shows an example of the many individual photos needed to develop classifier files for each feature to be detected.
These files are used to locate objects from within an image.
Next you can see the first iteration of the syringe detection software.
The entire frame is being processed leading to the slow frame rate.
False positives are also visible in the far upper left corner.
The improved feature detection limits the processed region to a small box in the center of the visible field.
This method reduces false positives while improving the processing time.
This improved processing is shown in the improved fame rate of the video.
The last demonstration is of a simple feature displacement algorithm.
The two features coordinates are compared to one another and the hypotenuse between the points is displayed and measured in pixels.
The next step would be to allow for accurate scaling of the image to compensate for camera distance and parallax.
