balloon61 / perception-project

Assignments in Computer Vision course. This repository including AR marker detect, Histogram equalization, Line detection, Stereo Vision, Turning prediction

Perception-Project

Description:

This repository includes several basic perception project

Requirement:

numpy, cv2, matplotlib

AR Tag Detection:

Description:

Detect AR tag in any direction.

Result:

Video 1. AR Tag detection

Coin Seperation

Description:

Calculate how many coins in this image.

Result:

Image processing is shown as follows, and the number of coins will be published in your terminal when you executed the code.

Fig 1. (Left) Preprocessing image (Right) After processing image

Historgram Equalization

Description:

Given a sequence of image and do Historgram Equalization and Adaptive Historgram Equalization to each images.

Result:

Fig 2. Original image Fig 3. Historgram Equalization (top) original (mid) Historgram Equalization (bot) Adaptive Historgram Equalization

Line Detection

Description:

Detect Lane in the highway.

Video 2. Line Detection

Steering Prediction

Description:

Calculate the curvature of the lane in the highway and predict the steering direction.

Results:

Video 3. Turning Prediction

Stereo Vision

Description:

implement the concept of Stereo Vision. given several different datasets, each of them contains 2 images of the same scenario but taken from two different camera angles. By comparing the information about a scene from 2 vantage points, we can obtain the 3D information by examining the relative positions of objects.

Pipeline for creating a Stereo Vision System:

1. Calibration: Get the F (fundamental matrix), K (intrinsic matrix), and E (essential matrix)
2. Rectification: Apply the perspective transformation to make sure that the epipolar lines are horizontal for both images
3. Correspondence: Get the disparity image which is the image that given where each pixels gives the disparity of the 3D points.
4. Compute Depth Image: Directly use the disparity matrix to calculate the depth image.

Results:

Rectification:

Fig 4. Corresponding Eplliplolar line

Fig 5. Rectified

Correspondence:

Fig 6. Disparity

Depth Image:

Fig 7. Depth image