Peter Bardenhagen的动态

Recognizing the distortions in BWC and how it affects the hindsight findings; Recreating a crime scene through the lens of a 50mm camera provides crucial insights into spatial relationships, time, and motion. Our recent project highlighted how different lenses impact perception. Join for the Force Analysis; Video review and examination course in Chandler AZ, August 20-22, download the BWC report by CIR here https://lnkd.in/gnpDpPai #CrimeSceneRecreation #SpatialAnalysis

Inspired by Dr. Abdullah Al Redwan Newaz I was thinking of playing with Grounding DINO. I found the Grounding DINO has awesome features. To detect a new type of object in the image, we do not need to retrain our model.?The Grounding DINO has a prompt feature that detects new types of objects. For example, in an image, we are detecting persons. Next, we think about detecting the gun instead of the person. We just need to change our text prompt and then the model will detect the gun. To detect both the gun and the person we just need to change our text prompt accordingly. That's awesome!!!

Need to detect bats at long range? Maybe a railway tunnel project or a wind farm monitoring project? What kit do you need… We are at times limited by access restriction’s for site work, and when that happens we are still expected to provide our clients with hard evidence of presence/ absence or habitat use. Now we have access to thermal equipment for long range detection! Check out the clip below for a 200m + survey … ?????? If you want to know more get in touch with Thermal Vision Ecology #thermalimaging #remotesensing #batsurvey

Mini humanoid manages to kick a rolling ball B-Human conduct testing for RoboCup SPL 2025 which will take place from July 15 to 21, 2025 in Salvador, Brasil. B-Human is a collab between the Department of Computer Science of the University of Bremen and the DFKI research area Cyber-Physical Systems.

Acoustic Location: Is sound spatial and can you detect location through hearing? The short answer is yes, you can determine an object's location through sound. Examples range from its first use in World War I to modern day gunshot detection systems. https://lnkd.in/gDYuSgiq #locationintelligence

I'm so glad the holidays are over—me and vacation aren’t exactly the best of friends. It’s a joy to get back to work and start preparing a test plan, procedures, use cases, and timeline for some exciting new hardware we’re planning to add to the Knut system. The goal? To see how these upgrades integrate and perform alongside Munin, our Collaborative Combat Intelligence system. Here’s what’s on the testing list: A wide-angle camera (with a twist) to boost situational awareness. A new radiolink for more reliable communication. A laser seeker to sharpen precision targeting. This isn’t just about Knut. If all goes well, these features will pave the way for advancements across other projects, opening up some great possibilities. Full testing is scheduled for Q3, and I’m keeping my New Year’s resolution in mind: quicker iteration loops and wrapping up projects at “good enough”.

In the wake of our #NatureCommunications paper on wind field estimation aboard drones (https://lnkd.in/gxiNnPda), our #ICRA2024 work, "Watching the Air Rise" (https://lnkd.in/gz8qmeNn) takes another step towards enabling drones to model small-scale atmospheric phenomena. This time, it's about thermal updrafts, or thermals for short. Thermals are columns of rising air that extend from the ground up into the sky, a byproduct of atmospheric convection. They can help a small fixed-wing UAV gain altitude, and hence energy, "for free" -- if only the UAV can find them. Birds do so by following insects that get sucked into these air columns. Human glider pilots pinpoint some thermals using indirect evidence like cumulus clouds on top of them. At the lower altitudes where UAVs fly, though, most thermals don't have the cloud "markers" on top. Currently, UAVs rely on flying into a thermal by chance, since air movement is invisible to the cameras a drone may carry onboard. Or is it really? In the #ICRA2024 paper, we propose a method that allows a camera to detect the air rising in front of it from a *single* videoframe captured by a single *moving* optical camera. Previous attempts at doing this required a static multicamera setup -- impressive in its own right but impractical for putting onboard a small UAV. Despite current limitations due to background generalization, the progress that our approach, DYBS, achieves compared to the previous state of the art takes us closer to deploying an evolution of this method on drones one day! Come talk to us at #ICRA2024: May 14, 16:30-18:00 Tokyo time, Paper TuCT11-CC.3! ?? below ??

Modern Adversaries and Evasion Techniques. Why Legacy AV is an Easy Target. Watch on-demand here: https://oal.lu/NeBM8

Recapping Top Posts of 2024: Recognizing the distortions in BWC and how it affects the hindsight findings; Recreating a crime scene through the lens of a 50mm camera provides crucial insights into spatial relationships, time, and motion. Our recent project highlighted how different lenses impact perception. Join the Force Analysis; Video review and examination course in Chandler AZ, August 20-22, download the BWC report by CIR here https://buff.ly/41zEOat #CrimeSceneRecreation #SpatialAnalysis #BWC

We present a new article concerning one of our research fields: "Shock-tube tests on conventional windows: Exploring retrofit concepts for enhanced blast protection". The work was done in cooperation with Universit?t der Bundeswehr München and Technologie Campus Neustadt an der Donau. The full article ist available: https://lnkd.in/g2hqvuHN Keywords: #Experiment, #Protection, #Retrofitting, #Blast, #Shock_Tube, #Glas_Fiber_Reinforced_Polymer