dileep92s/studyproject

learn computer vision with opencv and python

studyproject

learn computer vision with opencv and python

studyproject's Issues

Detect lane when road and shoulder beside lane have two many white and yellow patches

use probabilistic hough line transform

probabilistic hough line transform is faster
it has an option to ignore small lines
it directly returns the points in XY plane, less calculation needed

stabilize lane overlay

draw deviation from center

assuming the camera is mounted on center, find out the deviation from center

Lane keeping with PID control

Lane keeping with PID control

store the output video

store the final output as output.mp4
this can be used to check current progress

trim the overlays

overlay is shown on full image as of now, trim it to ROI

game - if 2nd lane is not detected, stay away from detected lane

this is to avoid lane change,

Steering control using CNN

Lane Detection

Add code for lane detection.

Use threads for keypress

Lane Keeping with bang bang control

Add code for lane keeping.
First lane keeping with bang bang control

curved lane detection

optimize lane detection on curved roads

PID for steering control at all speed

apply HSL filter

apply HSL filter to detect only white and yellow lines
HSL was found to be more effective as per my study

don't detect horizontal lines

restructure the code

create a class
make the code modular

Detect all types of lane

lanes in udacity autonomous car simulator is different in compare to normal lanes, should be able to detect that kind of lanes robustly too

lane detection when lanes are partially covered with objects like leaves or fine particles

At the end it should be able to detect lane on udacity autonomous car simulator in autonomous mode

apply PID for steering control

@vijendra1125 don't know how to test, you don't have the game