Vehicle Detection Project
The goals / steps of this project are the following:
- Perform a Histogram of Oriented Gradients (HOG) feature extraction on a labeled training set of images and train a classifier Linear SVM classifier
- Optionally, you can also apply a color transform and append binned color features, as well as histograms of color, to your HOG feature vector.
- Implement a sliding-window technique and use your trained classifier to search for vehicles in images.
- Run your pipeline on a video stream (start with the test_video.mp4 and later implement on full project_video.mp4) and create a heat map of recurring detections frame by frame to reject outliers and follow detected vehicles.
- Estimate a bounding box for vehicles detected.
Rubric Points
###Here I will consider the rubric points individually and describe how I addressed each point in my implementation.
###Writeup / README
####1. Provide a Writeup / README that includes all the rubric points and how you addressed each one. You can submit your writeup as markdown or pdf. Here is a template writeup for this project you can use as a guide and a starting point.
You're reading it!
###Histogram of Oriented Gradients (HOG)
####1. Explain how (and identify where in your code) you extracted HOG features from the training images.
The code for this step is contained in the second code cell of the IPython notebook example.ipynb
I started by reading in all the vehicle and non-vehicle images. Here is an example of one of each of the vehicle and non-vehicle classes:
I then explored different color spaces and different skimage.hog() parameters (orientations, pixels_per_cell, and cells_per_block). I grabbed random images from each of the two classes and displayed them to get a feel for what the skimage.hog() output looks like.
Here is an example using the YCrCb color space and HOG parameters of orientations=8, pixels_per_cell=(8, 8) and cells_per_block=(2, 2):
####2. Explain how you settled on your final choice of HOG parameters.
I tried RGB, HSV, LUV and YCrCb color space. I also experimented with extracting HOG feature from single channel vs ALL channels. I changed other HOG parameters like pixels_per_cell and cells_per_block. I randomly selected 500 images and split them into test and train test. I then experimented with different paramerts and colro space and added histogram and spatial binning to figure out best combination. I then tried set of combinations that gave good result on full training set. I selected parameters which provided best accuracy on test set.
####3. Describe how (and identify where in your code) you trained a classifier using your selected HOG features and color features.
I trained a linear SVM using HOG, histogram and spatial binning features in YCrCb color space. The code the same in 3rd code cell in IPython notebook. I used train_test_split and shuffle functions to randomize data and split it into test and training sets. I also normalized features.
###Sliding Window Search
####1. Describe how (and identify where in your code) you implemented a sliding window search. How did you decide what scales to search and how much to overlap windows?
I started with experimenting on different window sizes from 32 to 256 in steps of 32 on test images. I also tried variable combinatio from 32 to 96 in variable steps of 12,16, 24. Evetually i fou dwindow sizes which gave best result. These were 64 to 160 in steps of 32. I also experimented with overlap values and saw 0.8 giving best results.
####2. Show some examples of test images to demonstrate how your pipeline is working.
Ultimately I reduced region of search by experimenting with x_start, x_stop, y_start, y_stop values, which provided a nice result. Here are some example images:
####1. Provide a link to your final video output. Your pipeline should perform reasonably well on the entire project video (somewhat wobbly or unstable bounding boxes are ok as long as you are identifying the vehicles most of the time with minimal false positives.) Here's a link to my video result
####2. Describe how you implemented some kind of filter for false positives and some method for combining overlapping bounding boxes.
I recorded the positions of positive detections in each frame of the video. From the positive detections I created a heatmap and then thresholded that map to identify vehicle positions. I then used scipy.ndimage.measurements.label() to identify individual blobs in the heatmap. I then assumed each blob corresponded to a vehicle. I constructed bounding boxes to cover the area of each blob detected.
Here's an example result showing the heatmap from a series of frames of video, the result of scipy.ndimage.measurements.label() and the bounding boxes then overlaid on the last frame of video:
Here is the output of scipy.ndimage.measurements.label() on the integrated heatmap from all six frames:
###Discussion
####1. Briefly discuss any problems / issues you faced in your implementation of this project. Where will your pipeline likely fail? What could you do to make it more robust?
I think manually choosing paramerts is tedious job, and i would consider deep learning techniques for vehicle detection if i have to do this again. Since my pipeline uses hand chosen features on small dataset, it may not generalize to light conditions and vehicle colors. I haven't added any logic to track detected vehicle. It would be nice to implement kalman filter to track detected vehicle even when it is occluded.