MASK R-CNN

The Mask R-CNN is a Convolution Neural Network algorithm, which was introduced by He et al. in their 2017 paper, Mask R-CNN. Mask R-CNN builds on the previous object detection work of R-CNN (2013), Fast R-CNN (2015), and Faster R-CNN (2015), all by Girshick et al.

My result is here:

OPENCV BOOK

In this folder is many tutorials for OpenCV. Recommend to take a look.

PRACTICAL CASES

In practical cases I am learning and dealing with practical cases with OpenCV such as:

• amazon cover search
• eye tracking
• face detection
• recognizing handwritten text
• object detection or plant classification

CENTER OF CONTOUR

goal: How do I compute the center of a contour using Python and OpenCV?

1. detect the outline of each shape in the image
2. compute the center of the contour — also called the centroid of the region

image preprocessing:

1. Conversion to grayscale
2. Blurring to reduce high frequency noise to make contour detection more accurate
3. Binarization of the image (thresholding)

SHAPE & COLOR DETECTION

This code is built on previos (center of contour) but here I also add something more. Algorithm is able to recognize shape and also color.

Detect shape:

1. Using Dougles-Peucker algorithm
2. peri = cv2.arcLength(contour, True)
3. approx = cv2.approxPolyDP(contour, 0.04*peri, True)
4. IF/ELSE based on number of detected lines (circle, triangle, square...)

TEXT CORRECTION

goal: find text and then rotate it to right - horizontal position for reading

1. Load image -> resize -> change colors into Gray
2. Bitwise_not -> background black, text white
3. Threshold (pixels will be just black and white)
4. Find coordinates and angle with: cv2.minAreaRect
5. Rotate image into horizontal position with: warpAffine based on: cv2.getRotationMatrix2D