OpenCV - Adaptive-thresholding / effective noise reduction?

Question

I am new to OpenCV and I just read about cv2.adaptiveThreshold() and decided to try it out. Sadly there is this huge amount of noise I cannot seem to get rid of.

What are some effective ways to reduce noise so I can draw proper contours? What is the best practice and why?

Here is the snippet:

import cv2
import numpy as np

#####################################
winWidth = 640
winHeight = 840
brightness = 100

cap = cv2.VideoCapture(0)
cap.set(3, winWidth)
cap.set(4, winHeight)
cap.set(10, brightness)

kernel = (5, 5)

###########################################

def preprocessing(frame):
    imgGray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
    # mask = cv2.inRange(imgHsv, lower, upper)
    imgBlurred = cv2.GaussianBlur(imgGray, kernel, 1)
    gaussC = cv2.adaptiveThreshold(imgBlurred, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY_INV, 11, 2)
    imgDial = cv2.dilate(gaussC, kernel, iterations=3)
    imgErode = cv2.erode(imgDial, kernel, iterations=1)

    return imgDial


def getcontours(imPrePro):
    contours, hierarchy = cv2.findContours(imPrePro, cv2.RETR_TREE, cv2.CHAIN_APPROX_NONE)
    for cnt in contours:
        cv2.drawContours(imgCon, cnt, -1, (255, 0, 0), 3)





###################################################

while (cap.isOpened()):
    success, frame = cap.read()
    if success == True:
        frame = cv2.flip(frame, 1)
        imgCon = frame.copy()
        imPrePro = preprocessing(frame)
        getcontours(imPrePro)
        cv2.imshow("Preprocessed", imPrePro)
        cv2.imshow("Original", imgCon)

        if cv2.waitKey(1) & 0xFF == ord("q"):
            break

Answer 1

I think it is best if we look at the blockSize and C parameters.

Form the source:

blockSize: Size of a pixel neighborhood that is used to calculate a threshold value for the pixel: 3, 5, 7, and so on.

C: Constant subtracted from the mean or weighted mean (see the details below). Normally, it is positive but may be zero or negative as well.

In your example you set C to 2:

gaussC = cv2.adaptiveThreshold(imgBlurred, 255, cv2.ADAPTIVE_THRESH_GAUSSIAN_C, cv2.THRESH_BINARY_INV, 11, 2)

C=2	C=8

As you can see, you need to play with blockSize and C parameters to get the desired result from adaptive-threshold.

In this question, we achieve less-noise by increasing C parameter.