aras62 / PIE

Annotations for Pedestrian Intention Estimation (PIE) dataset

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

What does the bounding box coordinates mean?

doMoreCode opened this issue · comments

commented

Hello, i dont understand the mean of bounding box coordinates such as xbr .Is that the object before the car?

commented

i just want to know where are the bounding box coordinates of pedestrian. @aras62 , thanks.

You can find the bounding box coordinates in attributes.zip for each pedestrian's instance, there is a 4-point coordinates like this: xbr="929.18" xtl="916.82" ybr="847.11" ytl="793.32" which correspond to top-left and bottom-right corners of the bounding boxes.

commented

You can find the bounding box coordinates in attributes.zip for each pedestrian's instance, there is a 4-point coordinates like this: xbr="929.18" xtl="916.82" ybr="847.11" ytl="793.32" which correspond to top-left and bottom-right corners of the bounding boxes.

emm i wanted to draw a rectangle to see which pedestrian in image, but i got none. In my case, i use matlab to manipulate image .(The origin of the picture is start from the upper left corner)。Should i use the camera's parameters to transfer the coordinates?@aras62

the coordinates in attributes are for opencv. Matlab has a different coordinate system, so you need to convert them. I don't remember matlab one but one of the main differences can be the axis. For your second question, no you do not need to you camera parameters, the coordinates are in image plane.

commented

the coordinates in attributes are for opencv. Matlab has a different coordinate system, so you need to convert them. I don't remember matlab one but one of the main differences can be the axis. For your second question, no you do not need to you camera parameters, the coordinates are in image plane.

thanks for your answer, i got it.