Color Filtering - OpenCV with Python for Image and Video Analysis 7

HSV values cannot be more than 100 for S and V.
I don't understand this notation. How can i create a range for my own color? For example i want a range for blue.

Is it possible to use python using a point cloud to extract only the road from the colors on both sides of the road (if there is a plantation like grass)?

can i like display the mask onto a jpg image ?

Check this course for an in-depth explanation of OpenCV from beginner to advanced: ​@9M38​

code ?

Use this for better red masking
lower_red = np.array([0,70,50])
upper_red = np.array([10,255,255])

i believe that it would be much easier if you use "Trackbars" instead of running the code many times.

I found this by searching ''whoops you have to put the cd in your computer''

I'think adjust value of mask by trackbars is a much better way.

Adam klavyeye bakmiyor ya adam sin adam

how can I change the colour of different part in same image,without disturbing the structure and rendering using any language that is either client side or server side.can u help me?

How can i get mean of the result so i can get how red is the hat? thanks for the videos

now I want a "Make computer vision great again!" hat

the value in HSV cannot be greater than 100 right? The V attribute tell how much the white light is ixed into the color. 100 means the color will be bright. 0 means it will be more like black

Just a quick note if you are as bad as me understanding color spaces:
Searching "opencv color picker" brought me to a couple of opencv scripts on GitHub that find the lower and upper values of any color by clicking a couple times the desired color on the image

My program is showing attribute error: module 'cv2.cv2' has no attribute'cvtcolor' please help me out

I have tried everything as it is but its not working for me. Can you help me out.

Thanks Foll All

Is there any algorithm for feature detection and matching , which considers the colors of the images.

I have this error
C:\Miniconda3\conda-bld\opencv-suite_1533128839831\work\modules\imgproc\src\color.cpp:11214: error: (-215) (scn == 3 || scn == 4) && (depth == 0 || depth == 5) in function cv::cvtColor


how to go about it?

so how do i run code when the color is detected?

Module 'cv2' has no 'COLOR_RGB2HSV' member

Adding a backtrack is so useful to adjust the colors in real time

Hey there mate
I actually want to detect a rgb trio and sort it by means of a servo
Plz send me some codings...

how can we change the color? ı didnt understand that . what if ı want make this with blue , how can i write code ?

i never enjoyed programming before watching your video..

Hey, I really like your videos. They help me a lot. What steps would you suggest, when trying to create a "building" recognition for ONE building? It should be able to detect parts of, so you can highlight lets say rooms from the outside.

What if you are using a kinect sensor? How can I interface it with that code? Thanks a lot

how can I detect hand/ skin color instead of red color of hat?
what values should I give in lower_red and upper_red?
here?
lower_red = np.array([150,50,50])
upper_red = np.array([180,255,255])

Is there any reason why you wrote k==27 ?

waitKey(0) doesn't display video. It displays a fixed image. The first image captured on the webcam.

𝐌𝐚𝐧𝐲 𝐩𝐞𝐨𝐩𝐥𝐞 𝐡𝐚𝐯𝐞 𝐝𝐨𝐮𝐛𝐭 𝐢𝐧 𝐭𝐡𝐢𝐬 𝐜𝐨𝐝𝐞:
𝐤 = 𝐜𝐯𝟐.𝐰𝐚𝐢𝐭𝐊𝐞𝐲(𝟓) & 𝟎𝐱𝐅𝐅
𝐢𝐟 𝐤 == 𝟐𝟕:
𝐛𝐫𝐞𝐚𝐤

𝟏. In cv2.waitKey(x) the parameter x indicates the delay in the frame display i.e a new frame is displayed in our video feed after x seconds (try keeping it 1000 to check the difference,1fps ).

𝟐. cv2.waitkey() returns the key pressed on the keyboard, according to the code above if ASCII value of that key is 27(ESC key) then we are breaking out of the loop.

𝟑. Now the 0xFF is a hexadecimal number = 255 which in binary is equal to 1111111 so doing a '&' Operation means we only care about the last 7 bits of the ASCII value returned.

Thx for the efforts and I have two questions :
1 - what are the details of
k = cv2.waitKey(5) & 0xFF
if k == 27:
break
and thier use
2 - how can we end the while loop and get destroy all windows to be executed

hi, i've got a problem, do you use the picamera to do this? because with my USB camera the pixel format is not 'BGR'

I've been wracking my brains around how these color masks work all day, thanks!

what is the range of sat and val in opencv ?

is the usual range in (0-100%)?

hsv = cv2.cvtColor(frame, cv2.COLOR_BGR2HSV)
error: ..\..\..\modules\imgproc\src\color.cpp:3961: error: (-215) (scn == 3 || scn == 4) && (depth == CV_8U || depth == CV_32F) in function cv::cvtColor

What does waitKey(5) do?

I don't know how to determine lower and higher HSV array?

why do you need the underscore in this line? _, frame = cap.read()

Just to add some more..... Red lies in a region where we cant define using single numpy-array/mask. So, I've done this!
## only for red -30 to 0 and 0 to 30 - 2masks
lower_red = np.array([150,50,50])
upper_red = np.array([180,255,255])
mask1 = cv2.inRange(hsv_version, lower_red, upper_red)

lower_red = np.array([0,50,50])
upper_red = np.array([30,255,255])
mask2 = cv2.inRange(hsv_version, lower_red, upper_red)
mask = cv2.bitwise_or(mask1,mask2)
res = cv2.bitwise_and(frame, frame, mask=mask)

Hey sentdex, Guys and Girls,

i want to recognize a color, and if it is recognized i want to count it or give it a value for example (red = 50, Black = 20).

if red or black was recognized I should be issued the number.

any Idea ????

need help :-)

Please explain in a bit more detail otherwise Awesome video

what camera did you use for this project?

So if I am using a video, python speeds up the video to about 3x or 4x the original speed. Can I slow that down somehow? Didn't find anything in the opencv documentation at first glance.

This is a good tutorial.But I need to mask green pixels .So can you pls mention the green color's upper bound and lower bound?