Just an idea: if you could try to establish a known colour or texture of the floor using computer vision then avoid anything that doesn't look like the floor. That's what I do when I walk cautiously. Probably the most robust floor detector for a "very unstructured environment" would be a CNN trained to segment floor vs not-floor. If you have the budget, one way to get a training dataset is to create one using amazon turk to manually label a lot of images. Here's a couple slides showing existing deep learning segmentation approaches and software that I think look promising. https://www.slideshare.net/danielsnider/deep-learning-segmentation Here's a list of deep learning segmentation publications: https://handong1587.github.io/deep_learning/2015/10/09/segmentation.html Another list of deep learning computer vision: https://github.com/kjw0612/awesome-deep-vision --- [Visit Topic](https://discourse.ros.org/t/avoiding-small-obstacles-on-the-gruond/1939/9) or reply to this email to respond. If you do not want to receive messages from ros-users please use the unsubscribe link below. If you use the one above, you will stop all of ros-users from receiving updates. ______________________________________________________________________________ ros-users mailing list ros-users@lists.ros.org http://lists.ros.org/mailman/listinfo/ros-users Unsubscribe: