Image Processing example and functions needed

Arduino SDK
1

The application is for Road Lane Detection, (part of AI cam module project)

OpenCV | Real Time Road Lane Detection - GeeksforGeeks

the needed mage processing functions are

  1. cv2.gray

  2. cv2.GaussianBlur(grayscale, (5,5), 0)

  3. cv2.Canny(gray_img, 50, 150)

  4. region selection

cv2.fillPoly(mask, vertices, ignore_mask_color)
masked_image = cv2.bitwise_and(image, mask)
  1. cv2.HoughLinesP(image, rho = rho, theta = theta, threshold = threshold,
    minLineLength = minLineLength, maxLineGap = maxLineGap)

Please support these image process function calls. Thanks !

2

To All,

I have tried AI agent coding, it can convert RGB to Gray, but failed at 5x5 Gaussian Blur

probably too many steps of processing on entire 640x480 image.

The alternative for road lane detection is to use YOLOv7 to detect the lane as a bounding box,

so the lane center can be identified.

3

Hi @rkuo2000 ,

we are currently reviewing this and will update you again. Thank you

1 Like
4

I have just done something similar using opencv.js on esp32, I tried to do it on an amb82-mini, but I couldn’t find a way of filtering the image. the the amb82 is much better in just about every way other than software support. it would be nice if opencv.js worked, its basically just serving a webpage so i dont see why it wouldnt as all the processing is done on the browser.

im using a modified version of this, it uses an rgb filter to track a colour range, the tracking range changes based on how wide the path it detects is and how close it gets to the edge of the screen. then tracks the centre of that mass and steers towards it.

1 Like