DeerDar: A Novel Application of Computer Vision for Road Safety

Photo of our car!

Bridging the gap between theoretical data science and tangible safety solutions, I introduce DeerDar – my novel contribution to the realm of computer vision and road safety. As a dedicated mathematician and data scientist, I've channeled my expertise into the development of this unique system, leveraging advanced computer vision techniques to detect wildlife and alert drivers. DeerDar is not just a device; it's the embodiment of my journey to address the real-world problem of road collisions with wildlife, one innovative detection at a time.

The Core of DeerDar: Where Mathematics Meets Machine

Photo of the Raspberry Pi, the computer component of DeerDar.

Every significant invention begins with a vision. For DeerDar, it started with this Raspberry Pi computer – a compact yet mighty brain capable of running complex algorithms. As a solo developer, I began this project by programming this device to process visual data, laying the groundwork for the sophisticated functionalities to come.

Designing DeerDar: A Blueprint for Safety

Photo of the CAD model of DeerDar that I constructed in Fusion 360 for this project.

With the core computing element in place, I turned to design. The CAD model you see above was my first step in bringing DeerDar to life. It represents hours of meticulous planning and precise mathematical calculations, ensuring that each curve and angle serves both form and function, a true manifestation of my background in mathematics and data science.

Crafting DeerDar: From Digital Design to Physical Form

Photos of the 3D printing process of DeerDar's casing.

The translation from digital blueprints to tangible reality is a transformative one. The images here capture the 3D printing process of DeerDar's casing. They represent a critical phase in the project where my digital designs materialized into the physical world, layer by precise layer. It's a testament to the DIY spirit and the possibilities that modern technology affords individual creators.

Bringing DeerDar to Life: Assembly and Completion

Photo of the complete DeerDar setup with the LCD screen. Photo of the front view of the DeerDar device with a mounted camera.

Bringing a concept into reality is a transformative journey, especially when it embodies a personal mission. The images above capture a milestone moment: the assembly of the DeerDar device. This is where the myriad pieces, borne from calculations and design, coalesce into a singular, functional entity.

The Innovative Vision System: DeerDar's Eyes

Photo of the DeerDar device with the infrared and standard camera setup.

Finally, we arrive at the operational prototype of DeerDar. The device, as depicted above, is outfitted with a day-and-night vision system, using both standard and infrared cameras. This dual-capability ensures round-the-clock monitoring and data capture, critical for detecting wildlife at any hour. The glowing LEDs are not just for aesthetics—they indicate the system's readiness to safeguard your journey.

DeerDar is more than just a project; it's a passion fueled by personal experience, a blend of data science, mathematics, and a touch of creativity. Currently in the production phase, the device aims to incorporate the YOLO (You Only Look Once) algorithm for real-time animal detection, a feature that promises to bring my vision of safer roads to fruition. As I continue to develop DeerDar, I am reminded that the journey of innovation is never traveled alone. It's a path shared with every future user whose drive will be a little safer, thanks to this technology.

Stay tuned for the journey ahead, where data science and practical application converge to create safer roads for all. On to the data collection phase!