Use in field biology
Good-day,
I am a field biologist trying to find a unit that would allow me to do two things:
1) Create 3D scans of mushrooms.
2) Census flying insects against the sky. I would like to use LiDAR (or similar technology) to get periodic, countable snapshots of insects flying above an upward-facing camera (i.e., with the sky as background). I would not expect to be able to identify the insects nor detect objects more than perhaps 4-5m away. However, ideally I would want to be able to automatically count the insects in front of the camera when the image was taken.
Would the L515 or some other unit work for such purposes?
Thank you,
Conrad.
-
Good-morning Marty,
So, now that Python talks with the camera, I am going back to my goal of insect surveying.
Basically, to restate that objective, I envision laying the camera on its back pointing upwards at the empty sky (or pointed towards some other distant background) and taking still shots at fixed intervals. I imagine insects (and sometimes birds) flying above it at various distances from the camera. My first simplistic goal is to take that shot and count all discrete objects registered within, say, 5m of the camera. Clearly, the smaller the object, the faster it is moving, and the farther it is from the camera, the lower my chance of capturing it. However, because my goal here is RELATIVE abundances, I'm OK, at least initially, if some objects are not 'caught' by the camera as long as I get semi-standardized data that I can compare over time.
I am not at this point hoping to ID the insects or detect their movement patterns. My guess is that the general work flow would be: take still photo with depth info (I do not need RGB info. and would hope to be able to do this day & night), generate image of those objects within say 5m of the camera, use some image counting function to count discrete blobs in camera. I imagine we might want to get more complicated eventually, but that would be a start.
Given this goal & reviewing your initial comments, it sounds as if the C++ routine you mention (https://github.com/IntelRealSense/librealsense/tree/master/examples/align-advanced) might be appropriate. I am not a programmer/developer so the more 'cook-book' the better. My questions are - is there now way of creating those images I want within the RealSense viewer or some other GUI? Is there a way of doing the same thing as the cpp routine from Python? If I try to run that cpp example, what interface do I do that from? (For example, I have been running Python from IDLE.) Would trying to get numby and/or openCV activated give me the ability to do what I want to do and provide some additional flexibility?
Just trying to get some general direction and then I try to progress, asking questions as I get stumped.
Thanks again.
-
If the camera is pointing upwards then it is likely that there will be nothing in the camera's view except for a series of blobs that represent insects. In that case, an OpenCV tutorial in the link below for detecting coins on an RGB image and counting them may be adaptable for your application.
https://medium.com/analytics-vidhya/detecting-and-counting-objects-with-opencv-b0f59bc1e111
-
Thank you. Yes, I think that could be part of the ingredients, except that I want to be able to use the depth function to remove birds (who will be larger but more distant than insects) based on a depth function &, ultimately, I might want to orient the camera horizontally which means there will be a distant landscape to be removed. And I want to be able to do this at night, meaning, I assume, the RGB will not be relevant (but, presumably, a similar approach to what you suggest could be taken with a non RGB image).
-
The link below provides Python resources for programming a depth Threshold Filter in the RealSense SDK to control the distance rendered on the depth image.
https://github.com/IntelRealSense/librealsense/issues/5964#issuecomment-594372943
Please sign in to leave a comment.
Comments
35 comments