Computer vision needle tracking through ballistic gel.

Hello,

I recently got help from the forum for tracking the depth and angle of a needle in air, however, now the camera is viewing the needle through gel, so there is quite a lot of noise, and the method that was previously working in air, doesn't work so well through gel. I would appreciate some help in getting the code to accurately detect the needle depth and angle through the gel, without picking up all of the surrounding noise instead. I have attached a .zip file with the code and relevant pictures.

Thanks.

Answer 1

[Image Analyst]

Honestly those some of those pictures are so bad that it will be difficult to accurately find the needle. You'll accomplish your task (measuring all the needles) faster if you just have your user manually draw a line along the needle using drawline. I'm attaching an example. Even though you have to manually point to the end(s) of the needle it will be faster than trying to develop an algorithm to do it automatically.

[Bozza]

Yeah, I thought that may have been the case. I'll try some hardware and lighting changes to see if I can improve the quality. I can't draw the lines manually as the aim is to have this working in real-time with the needle in varying positions.