On the established 2D map, the positions of the targets are marked. The targets are labeled with points in accordance with this sequence” rack , cushion, lamp, stair cooktop

We also store the 2d map call “map.py”

python build2dmap.py

---

python main.py -p map.png -s 80

In this part you can choose the stepsize the 2d map you like to use.

Next, the system will ask you about your target. Afterwards, you will be presented with a 2D map to select your starting position.

We also output path point called “movingoath.npy”

Below is the result of RRT algorithm implementation

• cooktop

• cushion

• lamp

• rack

• stair

---

python load.py -p <target>

We highlight the target with a transparent mask while navigating, so that we can clearly visualize the target category