Ethan Wellenreiter
423b511dd9
Moving around the testing notebooks. Autocropping is about done with exception to any new versions or converting the stuff to C code. Signed-off-by: Ethan Wellenreiter <ewellenreiter@gmail.com>
317 lines
9.1 KiB
Plaintext
317 lines
9.1 KiB
Plaintext
{
|
|
"cells": [
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# can probably be deleted or put somewhere. Was the original code for the rowsumdeskew"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 36,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"import cv2\n",
|
|
"import numpy as np\n"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 37,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"\n",
|
|
"\n",
|
|
"src = 255 - cv2.imread('./testing_space/cropped1.jpg',0)\n",
|
|
"scores = []\n",
|
|
"\n",
|
|
"h,w = src.shape\n",
|
|
"small_dimention = min(h,w)\n",
|
|
"src = src[:small_dimention, :small_dimention]\n",
|
|
"\n",
|
|
"out = cv2.VideoWriter('./temp/video.avi',\n",
|
|
" cv2.VideoWriter_fourcc('M','J','P','G'),\n",
|
|
" 15, (320,320))\n",
|
|
"\n",
|
|
"src = cv2.threshold(src, 100, 255, cv2.THRESH_BINARY)[1]"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 38,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"def rotate(img, angle):\n",
|
|
" rows,cols = img.shape\n",
|
|
" M = cv2.getRotationMatrix2D((cols/2,rows/2),angle,1)\n",
|
|
" dst = cv2.warpAffine(img,M,(cols,rows))\n",
|
|
" return dst\n",
|
|
"\n",
|
|
"def sum_rows(img):\n",
|
|
" # Create a list to store the row sums\n",
|
|
" row_sums = []\n",
|
|
" # Iterate through the rows\n",
|
|
" for r in range(img.shape[0]-1):\n",
|
|
" # Sum the row\n",
|
|
" row_sum = sum(sum(img[r:r+1,:]))\n",
|
|
" # Add the sum to the list\n",
|
|
" row_sums.append(row_sum)\n",
|
|
" # Normalize range to (0,255)\n",
|
|
" row_sums = (row_sums/max(row_sums)) * 255\n",
|
|
" # Return\n",
|
|
" return row_sums\n",
|
|
"\n",
|
|
"def display_data(roi, row_sums, buffer): \n",
|
|
" # Create background to draw transform on\n",
|
|
" bg = np.zeros((buffer*2, buffer*2), np.uint8) \n",
|
|
" # Iterate through the rows and draw on the background\n",
|
|
" for row in range(roi.shape[0]-1):\n",
|
|
" row_sum = row_sums[row]\n",
|
|
" bg[row:row+1, :] = row_sum\n",
|
|
" left_side = int(buffer/3)\n",
|
|
" bg[:, left_side:] = roi[:,left_side:] \n",
|
|
" cv2.imshow('bg1', bg)\n",
|
|
" k = cv2.waitKey(1)\n",
|
|
" out.write(cv2.cvtColor(cv2.resize(bg, (320,320)), cv2.COLOR_GRAY2BGR))\n",
|
|
" return k\n"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 39,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"count = 0\n",
|
|
"othercount = 0\n",
|
|
"goodangle = 0"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 40,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"# cv2.imshow('bg1', src)\n",
|
|
"# cv2.waitKey(0)\n",
|
|
"# cv2.destroyAllWindows()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 41,
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"found optimal rotation\n"
|
|
]
|
|
},
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n",
|
|
"found optimal rotation\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"# Rotate the image around in a circle\n",
|
|
"angle = 0\n",
|
|
"while angle <= 360:\n",
|
|
" # Rotate the source image\n",
|
|
" img = rotate(src, angle) \n",
|
|
" # Crop the center 1/3rd of the image (roi is filled with text)\n",
|
|
" h,w = img.shape\n",
|
|
" buffer = min(h, w) - int(min(h,w)/1.5)\n",
|
|
" roi = img[int(h/2-buffer):int(h/2+buffer), int(w/2-buffer):int(w/2+buffer)]\n",
|
|
" # Create background to draw transform on\n",
|
|
" bg = np.zeros((buffer*2, buffer*2), np.uint8)\n",
|
|
" # Compute the sums of the rows\n",
|
|
" row_sums = sum_rows(roi)\n",
|
|
" # High score --> Zebra stripes\n",
|
|
" score = np.count_nonzero(row_sums)\n",
|
|
" scores.append(score)\n",
|
|
" othercount = othercount + 1\n",
|
|
" # Image has best rotation\n",
|
|
" if score <= min(scores):\n",
|
|
" count = count + 1\n",
|
|
" # Save the rotatied image\n",
|
|
" print('found optimal rotation')\n",
|
|
" best_rotation = img.copy()\n",
|
|
" goodangle = angle\n",
|
|
" k = display_data(roi, row_sums, buffer)\n",
|
|
" if k == 27: break\n",
|
|
" # Increment angle and try again\n",
|
|
" angle += .75\n",
|
|
"cv2.destroyAllWindows()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": 42,
|
|
"metadata": {},
|
|
"outputs": [
|
|
{
|
|
"name": "stdout",
|
|
"output_type": "stream",
|
|
"text": [
|
|
"25\n",
|
|
"481\n",
|
|
"349.5\n"
|
|
]
|
|
}
|
|
],
|
|
"source": [
|
|
"print(count)\n",
|
|
"print(othercount)\n",
|
|
"print(goodangle)\n",
|
|
"cv2.imshow('bg1', best_rotation)\n",
|
|
"cv2.waitKey(0)\n",
|
|
"cv2.destroyAllWindows()"
|
|
]
|
|
},
|
|
{
|
|
"cell_type": "code",
|
|
"execution_count": null,
|
|
"metadata": {},
|
|
"outputs": [],
|
|
"source": [
|
|
"print(\"start\")\n",
|
|
"\n",
|
|
"# Rotate the image around in a circle\n",
|
|
"angle = 0\n",
|
|
"while angle <= 360:\n",
|
|
" # Rotate the source image\n",
|
|
" img = rotate(src, angle) \n",
|
|
" # Crop the center 1/3rd of the image (roi is filled with text)\n",
|
|
" h,w = img.shape\n",
|
|
" buffer = min(h, w) - int(min(h,w)/1.5)\n",
|
|
" #roi = img.copy()\n",
|
|
" roi = img[int(h/2-buffer):int(h/2+buffer), int(w/2-buffer):int(w/2+buffer)]\n",
|
|
" # Create background to draw transform on\n",
|
|
" bg = np.zeros((buffer*2, buffer*2), np.uint8)\n",
|
|
" # Threshold image\n",
|
|
" _, roi = cv2.threshold(roi, 140, 255, cv2.THRESH_BINARY)\n",
|
|
" # Compute the sums of the rows\n",
|
|
" row_sums = sum_rows(roi)\n",
|
|
" # High score --> Zebra stripes\n",
|
|
" score = np.count_nonzero(row_sums)\n",
|
|
" if sum(row_sums) < 100000: scores.append(angle)\n",
|
|
" k = display_data(roi, row_sums, buffer)\n",
|
|
" if k == 27: break\n",
|
|
" # Increment angle and try again\n",
|
|
" angle += .5\n",
|
|
" print(\"loop\")\n",
|
|
"cv2.destroyAllWindows()\n",
|
|
"\n",
|
|
"print(\"endofrotate\")\n",
|
|
"\n",
|
|
"# Create images for display purposes\t\n",
|
|
"display = src.copy()\n",
|
|
"# Create an image that contains bins. \n",
|
|
"bins_image = np.zeros_like(display)\n",
|
|
"for angle in scores:\n",
|
|
" # Rotate the image and draw a line on it\n",
|
|
" display = rotate(display, angle) \n",
|
|
" cv2.line(display, (0,int(h/2)), (w,int(h/2)), 255, 1)\n",
|
|
" display = rotate(display, -angle)\n",
|
|
" # Rotate the bins image\n",
|
|
" bins_image = rotate(bins_image, angle)\n",
|
|
" # Draw a line on a temporary image\n",
|
|
" temp = np.zeros_like(bins_image)\n",
|
|
" cv2.line(temp, (0,int(h/2)), (w,int(h/2)), 50, 1)\n",
|
|
" # 'Fill' up the bins\n",
|
|
" bins_image += temp\n",
|
|
" bins_image = rotate(bins_image, -angle)\n",
|
|
" \n",
|
|
"print(\"endofbins\")\n",
|
|
"\n",
|
|
"# Find the most filled bin\n",
|
|
"for col in range(bins_image.shape[0]-1):\n",
|
|
"\tcolumn = bins_image[:, col:col+1]\n",
|
|
"\tif np.amax(column) == np.amax(bins_image): x = col\n",
|
|
"for col in range(bins_image.shape[0]-1):\n",
|
|
"\tcolumn = bins_image[:, col:col+1]\n",
|
|
"\tif np.amax(column) == np.amax(bins_image): y = col\n",
|
|
"# Draw circles showing the most filled bin\n",
|
|
"cv2.circle(display, (x,y), 560, 255, 5)\n",
|
|
"\n",
|
|
"print(\"plotting\")\n",
|
|
"\n",
|
|
"# Plot with Matplotlib\n",
|
|
"import matplotlib.pyplot as plt\n",
|
|
"import matplotlib.image as mpimg\n",
|
|
"f, axarr = plt.subplots(1,3, sharex=True)\n",
|
|
"axarr[0].imshow(src)\n",
|
|
"axarr[1].imshow(display)\n",
|
|
"axarr[2].imshow(bins_image)\n",
|
|
"axarr[0].set_title('Source Image')\n",
|
|
"axarr[1].set_title('Output')\n",
|
|
"axarr[2].set_title('Bins Image')\n",
|
|
"axarr[0].axis('off')\n",
|
|
"axarr[1].axis('off')\n",
|
|
"axarr[2].axis('off')\n",
|
|
"plt.show()\n",
|
|
"\n",
|
|
"cv2.waitKey()\n",
|
|
"cv2.destroyAllWindows()"
|
|
]
|
|
}
|
|
],
|
|
"metadata": {
|
|
"kernelspec": {
|
|
"display_name": "Python 3",
|
|
"language": "python",
|
|
"name": "python3"
|
|
},
|
|
"language_info": {
|
|
"codemirror_mode": {
|
|
"name": "ipython",
|
|
"version": 3
|
|
},
|
|
"file_extension": ".py",
|
|
"mimetype": "text/x-python",
|
|
"name": "python",
|
|
"nbconvert_exporter": "python",
|
|
"pygments_lexer": "ipython3",
|
|
"version": "3.10.12"
|
|
},
|
|
"orig_nbformat": 4
|
|
},
|
|
"nbformat": 4,
|
|
"nbformat_minor": 2
|
|
}
|