projecting side checkard board onto bed of CNC seems to be working, need to make it work for other side of cnc bed, then use those results to generate bed homography

realWorldGcodeSender.py

@@ -255,6 +255,43 @@ def onclick(event):
   overlay = overlaySvg2(output, pixelToMmTransformMatrix, newPointIn[0][0], newPointIn[1][0])
   matPlotImage.set_data(cv2.cvtColor(overlay, cv2.COLOR_BGR2RGB))
 
+def crop_half_vertically(img):
+  #cropped_img = image[,int(image.shape[1]/2):int(image.shape[1])]
+  #height = img.shape[0]
+  width = img.shape[1]
+  # Cut the image in half
+  width_cutoff = int(width // 2)
+  left = img[:, :width_cutoff]
+  right = img[:, width_cutoff:]
+  return left, right
+
+def generate_dest_locations(boxWidth, corners, image):
+  prevX=0
+  prevY=0
+  locations = []
+  yIndex = 0
+  xIndex = 0
+  for corner in corners:
+    x,y= corner[0]
+    x= int(x)
+    y= int(y)
+
+    #cv2.rectangle(gray, (prevX,prevY),(x,y),(i*3,0,0),-1)
+    image = cv2.arrowedLine(image, (prevX,prevY), (x,y),
+                                     (255,255,255), 5)
+    locations.append([xIndex * boxWidth, yIndex * boxWidth])
+    if xIndex == 2:
+      xIndex = 0
+      yIndex = yIndex + 1
+    else:
+      xIndex = xIndex + 1
+    prevX = x
+    prevY = y
+  return locations, image
+
+def pixel_loc_at_cnc_bed(boxWidth, backward):
+  return cv2.perspectiveTransform(np.array([[0,0],[boxWidth*8,0],[boxWidth * 8,boxWidth*22],[0,boxWidth*22]]).reshape(-1,1,2), backward)
+
 #############################################################################
 # Main
 #############################################################################
@@ -275,7 +312,8 @@ contours = []
 while len(contours) != 8:
   # Capture frame-by-frame
   #ret, frame = cap.read()
-  frame = cv2.imread('cnc.jpeg')
+  frame = cv2.imread('IMG_20211208_194811847.jpg')
+  img = cv2.imread('IMG_20211208_194811847.jpg')
 
   # load the image, clone it for output, and then convert it to grayscale
       
@@ -284,17 +322,50 @@ while len(contours) != 8:
   #######################################################################
   # Get grayscale image above threshold
   #######################################################################
-  mask = cv2.inRange(frame, (150, 150, 150), (255, 255, 255))
-  frame = cv2.bitwise_and(frame, frame, mask=mask)
+  #mask = cv2.inRange(frame, (150, 150, 150), (255, 255, 255))
+  #frame = cv2.bitwise_and(frame, frame, mask=mask)
   gray = cv2.cvtColor(frame, cv2.COLOR_BGR2GRAY)
+  gray1,gray2 = crop_half_vertically(gray)
+  gray = gray1
+  ret, corners = cv2.findChessboardCorners(gray, (3, 22), None)
+  boxWidth = 32.8125
+
+  #Generate destination locations
+  locations, gray = generate_dest_locations(boxWidth, corners, gray)
+
+  #Determine forward and backware transformation through homography
+  forward, status = cv2.findHomography(np.array(corners), np.array(locations))
+  backward, status = cv2.findHomography(np.array(locations), np.array(corners))
+  
+  im_out = cv2.warpPerspective(gray, forward, (int(800), int(1200)))
+  cv2.rectangle(im_out, (int(32.8125*7),int(32.8125*1)),(int(32.8125*8),int(32.8125*2)),(255,0,0),-1)
+
+  pixelsAtBed = pixel_loc_at_cnc_bed(boxWidth, backward)
+  #out = cv2.perspectiveTransform(np.array([[32.8125*7,32.8125*1],[32.8125*8,32.8125*1],[32.8125*8,32.8125*2],[32.8125*7,32.8125*2]]).reshape(-1,1,2), backward)
+  line1 = tuple(pixelsAtBed[0][0].astype(np.int))
+  line2   = tuple(pixelsAtBed[1][0].astype(np.int))
+  line3   = tuple(pixelsAtBed[2][0].astype(np.int))
+  line4   = tuple(pixelsAtBed[3][0].astype(np.int))
+  cv2.line(gray, line1,line2,(255,0,0),3)
+  cv2.line(gray, line2,line3,(255,0,0),3)
+  cv2.line(gray, line3,line4,(255,0,0),3)
+  cv2.line(gray, line4,line1,(255,0,0),3)
+
+  
+  cv2.imshow('dst',im_out)
+  gray = cv2.resize(gray, (1280, 800))
+  
+  cv2.imshow('image',gray)
+  cv2.waitKey()
   #gray = blackWhite
+  print(gray.shape)
   
   #######################################################################
   # filter out noise
   #######################################################################
   # apply GuassianBlur to reduce noise. medianBlur is also added for smoothening, reducing noise.
-  gray = cv2.GaussianBlur(gray,(7,7),0);
-  gray = cv2.medianBlur(gray,7)
+  #gray = cv2.GaussianBlur(gray,(7,7),0);
+  #gray = cv2.medianBlur(gray,7)
   
   #######################################################################
   # Get black and white image of gray scale
@@ -306,12 +377,12 @@ while len(contours) != 8:
   # Find edges
   #######################################################################
   #Adaptive Guassian Threshold is to detect sharp edges in the Image. For more information Google it.
-  gray = cv2.adaptiveThreshold(gray,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
-            cv2.THRESH_BINARY,9,10.5)
+  #gray = cv2.adaptiveThreshold(gray,255,cv2.ADAPTIVE_THRESH_GAUSSIAN_C,\
+  #          cv2.THRESH_BINARY,9,10.5)
   
-  kernel = np.ones((5,5),np.uint8)
-  gray = cv2.erode(gray,kernel,iterations = 1)
-  gray = cv2.dilate(gray,kernel,iterations = 1)
+  #kernel = np.ones((5,5),np.uint8)
+  #gray = cv2.erode(gray,kernel,iterations = 1)
+  #gray = cv2.dilate(gray,kernel,iterations = 1)
 
   
   #######################################################################
@@ -363,6 +434,9 @@ while len(contours) != 8:
     blackWhite = cv2.bitwise_and(blackWhite, blackWhite, mask=circleMask)
     contours, hierarchy = cv2.findContours(blackWhite,cv2.RETR_TREE,cv2.CHAIN_APPROX_NONE)
     #print("Contours found: " + str(len(contours)))
+    cv2.imshow('image',gray)
+    cv2.waitKey()
+    
 
      
 #######################################################################