I am mapping rooms and corridors to be visualized on mobile application. As shown in the picture the map consists of two layers of binary data one for the occupied area (the red one) and one for not occupied areas (the green one), and not discovered area (the black area).
In this map the occupied areas is surrounding the not occupied but that is not the case all time, occupied areas could be surrounded by not occupied one.
The mapping process run on closed rooms where multi uniform and non uniform shaped objects presents. And formed out of 3D visual scanning device transformed to 2D occupancy map. Meaning that it’s not possible to maintain data to form a solid shapes like 2D Lidar scanning. There is no need to classify the objects, the only need to enhance the map to be good to visual on application. Map defined as two numpy Boolean arrays with variant dimension between 1000x1000 to 5000x5000 (not square shaped all time). The most important criteria is the processing time; I am processing the data on NVIDIA Jetson nano board, so processing power and time need to be minimized.
The processing time of map with 1000x1000 cannot exceed 300 ms. My main code is in Python, but for the speed the process up I prefer the enhancing code to be built in c++ (CUDA if possible) with Python to run it from python.
Here is examples of the needed output of demanded program: Also map need to be aligned vertically. The following image is the output needed for both layers (Image 2);
please check the attachment
14 фрилансеров(-а) готовы выполнить эту работу в среднем за $596
Hi, We can help you for Map enhancement for visualization. So sizes will vary from 1000 to 5000? Please confirm on this. Best Regards Rahul Kumar
Hi - This job looks like a good fit with my skill set and experience. I hold Bachelor of Computer Science and Master of Data Science Please see my profile and reviews for references.
Hey, I have checked your requirement and understand that as well. I have done SIMILAR work past. Do you want to see the DEMO WORK??? Will show you Thanks.