adal-glez / adv_lane_lines

Advanced Lane Finding Project for Self-Driving Car ND

Advanced Lane Finding Project

Adalberto Gonzalez

Self-Driving Car Engineer Nanodegree Program

The goals / steps of this project are the following:

• Compute the camera calibration matrix and distortion coefficients given a set of chessboard images.
• Apply a distortion correction to raw images.
• Use color transforms, gradients, etc., to create a thresholded binary image.
• Apply a perspective transform to rectify binary image ("birds-eye view").
• Detect lane pixels and fit to find the lane boundary.
• Determine the curvature of the lane and vehicle position with respect to center.
• Warp the detected lane boundaries back onto the original image.
• Output visual display of the lane boundaries and numerical estimation of lane curvature and vehicle position.

README

Camera Calibration

The code for this step is contained in the 3rd block cell of the IPython notebook located in this repository of the file called advanced-lane-lines.ipynb.

I start by preparing "object points", which will be the (x, y, z) coordinates of the chessboard corners in the world. Here I am assuming the chessboard is fixed on the (x, y) plane at z=0, such that the object points are the same for each calibration image. Thus, objp is just a replicated array of coordinates, and objpoints will be appended with a copy of it every time I successfully detect all chessboard corners in a test image. imgpoints will be appended with the (x, y) pixel position of each of the corners in the image plane with each successful chessboard detection.

I then used the output objpoints and imgpoints to compute the camera calibration and distortion coefficients using the cv2.calibrateCamera() function. I applied this distortion correction to the test image using the cv2.undistort() function and obtained this result:

Pipeline (single images)

To demonstrate this step, I will describe how I apply the distortion correction to one of the test images like this one:

the output from camera camera calibration objpoints and imgpoints is applied here using cv2.calibrateCamera() function. I applied this distortion correction to the test image using the cv2.undistort().

Color transforms.

I used a combination of color and gradient thresholds to generate a binary image (locate this at the Threshold image section of the file ). Here's an example of my process for this step.

and Here's an example of my process result for this step.

3. Perspective transform.

The code for my perspective transform includes a function called cal_perspective(), which appears in section Perspective Transform ( birds-eye view) of the file. The cal_perspective() function takes as inputs an undistorted image.

This resulted in the following source and destination points:

Source	Destination
490, 482	0, 0
810, 482	1280, 720
1250, 720	1250, 720
40, 702	40, 0

I verified that my perspective transform was working as expected by drawing the src and dst points onto a test image and its warped counterpart to verify that the lines appear parallel in the warped image.

4. Lane-line identification

In the Finding Lanes section of the file I use a portion of the image to get an approximation fo the lanes position using a histogram approach

and then I scan sectiond of the image in order to detect lane presence to convert those positions in a curve using a 2nd order polynomial to get soemthing like this:

5. Radius of curvature .

based on this amazing tutorial I applied the next formuala found in Finding Lanes to calculate curvature and

6. Result Example

I implemented this step in Finding Lanes section in the function cal_fill_lines(). Here is an example of my result on a test image:

---

Pipeline (video)

Here's a Link to my video result

---

Discussion

I have opportunities to improve deep dark road sections so i would like to explore L_channel combinations to improve that colection, i can also try a dynamic apporach on the length of the road since i have a fixed length and some roads have a tight curve and this apporach could need improvement i could also include a validation section to verify consistency among lane detections, so anything outside limits could be verified or discarded also i think i could try to work on the performance by going object oriented instead of lineal/structured.

As you could see this projet is far from being completed because even when i feel i completed the task, there is a enormous oportunity to growth and improvement, i hope i can come back to this project to improve on my thoughts on new learnings!

Thank you for reading this.

Adalberto