def ps1_7():
start_time = time.time()
img = cv2.imread("../input/ps1-input2.png")
gray_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
media_img = cv2.GaussianBlur(gray_img, (9, 9), 3) # smooth image
# 区域1
plt.subplot(121)
plt.title("Canny后的图像", fontdict=font)
edge_img = cv2.Canny(media_img, 30, 50)
plt.imshow(edge_img, cmap='gray')
# 区域2
plt.subplot(122)
plt.title("查找到的圆", fontdict=font)
all_b = find_circles(edge_img,
min_radius=15,
max_radius=40,
threshold=140,
nhood_size=12)
print("共找到了圆的数量：", len(all_b))
drawn_img_b = hough_circles_draw(img, all_b) # 画圆
cv2.imwrite("../output/ps1-7-a-1.png", drawn_img_b)
drawn_img_b = cv2.cvtColor(drawn_img_b, cv2.COLOR_BGR2RGB)
plt.imshow(drawn_img_b)
print('\033[F\r7) Time elapsed: %.2f' % (time.time() - start_time))
plt.show()
def run():
img = cv2.imread('../input/ps1-input2.png', cv2.IMREAD_GRAYSCALE)
img_smoothed = cv2.GaussianBlur(img, (31, 31), 3)
edges_smoothed = cv2.Canny(img_smoothed, 0, 80)
centers, rads = find_circles(edges_smoothed, (25, 50), 100)
img1 = np.copy(img)
hough_circles_draw(img1, '../output/ps1-7-a-1.png', np.int16(centers),
np.int16(rads))
def ps1_7():
start_time = time.time()
img = cv2.imread('input/ps1-input2.png')
gray_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
eroded_img = cv2.erode(gray_img, np.ones((5,5),np.uint8), 1)
smoothed_img = cv2.blur(eroded_img, (3,3))
edge_img = auto_canny(smoothed_img, 0.5)
centers, radii = find_circles(edge_img, [20, 40], threshold=135, nhood_size=10)
img_circles = img.copy()
for i in range(len(radii)):
img_circles = hough_circles_draw(img_circles, 'output/ps1-7-a-1.png',
centers[i], radii[i])
print('\033[F\r7) Time elapsed: %.2f'%(time.time()-start_time))
def ps1_5():
start_time = time.time()
# 5a: Load coin image, smooth, detect edges and calculate hough space
img = cv2.imread("../input/ps1-input1.png")
gray_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
smoothed_img = cv2.GaussianBlur(gray_img, (9, 9), 3) # smooth image
edge_img = cv2.Canny(smoothed_img, 30, 50)
cv2.imwrite("../output/ps1-5-a-1.png", smoothed_img)
cv2.imwrite("../output/ps1-5-a-2.png", edge_img)
# detect circles with radius = 20 and save image
all_a = hough_peaks_circle(edge_img, radius=20, threshold=140, nhood_size=12)
drawn_img_a = hough_circles_draw(img, all_a) # 画圆
cv2.imwrite("../output/ps1-5-a-3.png", drawn_img_a)
# 5b: detect circles in the range [20 50]
all_b = find_circles(edge_img, min_radius=20, max_radius=30, threshold=180, nhood_size=12)
drawn_img_b = hough_circles_draw(img, all_b) # 画圆
cv2.imwrite("../output/ps1-5-b-1.png", drawn_img_b)
print('\033[F\r5) Time elapsed: %.2f s' % (time.time() - start_time))
def Q5(inputPath, outputPath):
origin_img = cv2.imread(os.path.join(inputPath, "ps1-input1.png"))
img_smooth = cv2.GaussianBlur(cv2.cvtColor(origin_img, cv2.COLOR_BGR2GRAY),
(25, 25), 3)
cv2.imwrite(os.path.join(outputPath, "ps1-5-a-1.png"), img_smooth)
img_canny = canny_sigma(img_smooth, 0.8)
cv2.imwrite(os.path.join(outputPath, "ps1-5-a-2.png"), img_canny)
H_20 = hough_circles_acc(img_canny, 20)
peaks_20 = hough_peaks(H_20, 10, 140, 100)
pic_circle_20 = hough_circles_draw(
origin_img.copy(), os.path.join(outputPath, "ps1-5-a-3.png"), peaks_20,
20)
centers, radius = find_circles(img_canny, [20, 50],
threshold=130,
nhood_size=10)
img_circles = origin_img.copy()
for i in range(len(radius)):
img_circles = hough_circles_draw(
img_circles, os.path.join(outputPath, "ps1-5-b-1.png"), centers[i],
radius[i])
return origin_img, img_smooth, img_canny, pic_circle_20, img_circles
def ps1_8():
start_time = time.time()
img = cv2.imread('input/ps1-input3.png', cv2.IMREAD_COLOR)
smoothed_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
smoothed_img = cv2.erode(smoothed_img, np.ones((3, ) * 2, np.uint8), 1)
smoothed_img = cv2.GaussianBlur(smoothed_img, (3, ) * 2, 2)
edge_img = cv2.Canny(smoothed_img, 40, 80)
# Detect lines
H, thetas, rhos = hough_lines_acc(edge_img)
peaks = hough_peaks(H, numpeaks=40, threshold=105, nhood_size=40)
peaks = filter_lines(peaks, thetas, rhos, 3, 24)
img_hl = hough_lines_draw(img, 'output/ps1-8-a-1.png', peaks, rhos, thetas)
# Detect circles
centers, radii = find_circles(edge_img, [20, 40],
threshold=110,
nhood_size=50)
img_circles = img.copy()
for i in range(len(radii)):
img_hl = hough_circles_draw(img_hl, 'output/ps1-8-a-1.png', centers[i],
radii[i])
print('\033[F\r8) Time elapsed: %.2f s' % (time.time() - start_time))
def ps1_5():
start_time = time.time()
# 5a: Load coin image, smooth, detect edges and calculate hough space
img = cv2.imread('input/ps1-input1.png', cv2.IMREAD_COLOR)
gray_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
smoothed_img = cv2.GaussianBlur(gray_img, (11, 11), 3) # smooth image
edge_img = auto_canny(smoothed_img, 0.8)
cv2.imwrite('output/ps1-5-a-1.png', smoothed_img)
cv2.imwrite('output/ps1-5-a-2.png', edge_img)
# detect circles with radius = 20 and save image
H_20 = hough_circles_acc(edge_img, 20)
peaks = hough_peaks(H_20, numpeaks=10, threshold=140, nhood_size=100)
img_circles = img.copy()
img_circles = hough_circles_draw(img_circles, 'output/ps1-5-a-3.png',
peaks, 20)
# 5b: detect circles in the range [20 50]
centers, radii = find_circles(edge_img, [20, 50],
threshold=153,
nhood_size=10)
img_circles = img.copy()
for i in range(len(radii)):
img_circles = hough_circles_draw(img_circles, 'output/ps1-5-b-1.png',
centers[i], radii[i])
print('\033[F\r5) Time elapsed: %.2f s' % (time.time() - start_time))
def ps1_8():
start_time = time.time()
img = cv2.imread("../input/ps1-input3.png")
gray_img = cv2.cvtColor(img, cv2.COLOR_BGR2GRAY)
media_img = cv2.medianBlur(gray_img, 5)
'''转换为俯视图的形式'''
o = 400, 200
x = 50
y = 100 + 15
min = 0
src = np.float32([[398 - min, 272], [372 - min, 54], [592 - min, 270],
[504 - min, 50]])
dst = np.float32([[o[0] - x, o[1] + y], [o[0] - x, o[1] - y],
[o[0] + x, o[1] + y], [o[0] + x, o[1] - y]])
M1 = cv2.getPerspectiveTransform(src, dst)
M2 = cv2.getPerspectiveTransform(dst, src)
img_warped = cv2.warpPerspective(media_img, M1, (650, 500))
img_warped_edge = cv2.Canny(img_warped, 30, 50)
img_original = cv2.warpPerspective(img, M1, (650, 500))
# plt.imshow(edge_img, cmap='gray')
# 区域1
plt.subplot(221)
plt.title("将圆画在Canny后的俯视图上", fontdict=font)
all_1 = find_circles(img_warped_edge,
min_radius=15,
max_radius=35,
threshold=170,
nhood_size=10)
# all_1 = [[110, 437, 15], [140, 345, 15], [394, 430, 15], [402, 429, 15], [193, 382, 15], [385, 433, 15],
# [395, 416, 15],
# [376, 427, 15], [403, 420, 15], [250, 436, 15], [231, 161, 16], [111, 438, 16], [191, 383, 16],
# [232, 312, 16],
# [294, 335, 16], [259, 435, 16], [406, 435, 16], [389, 428, 16], [395, 427, 16], [251, 435, 16],
# [181, 269, 17],
# [164, 322, 17], [191, 384, 17], [252, 436, 17], [232, 160, 17], [405, 436, 17], [158, 238, 17],
# [379, 435, 17],
# [258, 438, 17], [387, 434, 17], [205, 423, 18], [254, 366, 18], [190, 385, 18], [404, 439, 18],
# [293, 333, 18],
# [169, 320, 18], [163, 323, 18], [234, 310, 18], [158, 237, 18], [257, 437, 18], [185, 268, 19],
# [292, 332, 19],
# [206, 424, 19], [253, 367, 19], [299, 328, 19], [168, 321, 19], [114, 269, 19], [162, 236, 19],
# [380, 414, 19],
# [255, 436, 19], [255, 437, 20], [184, 268, 20], [161, 237, 20], [291, 332, 20], [390, 435, 20],
# [298, 329, 20],
# [380, 415, 20], [404, 425, 20], [424, 397, 20], [263, 433, 20], [262, 433, 21], [274, 237, 21],
# [297, 330, 21],
# [390, 437, 21], [407, 425, 21], [394, 422, 21], [413, 418, 21], [383, 448, 21], [400, 432, 21],
# [254, 437, 21],
# [383, 438, 22], [407, 427, 22], [413, 419, 22], [397, 435, 22], [387, 445, 22], [391, 418, 22],
# [391, 437, 22],
# [296, 329, 22], [381, 417, 22], [421, 417, 22], [408, 427, 23], [399, 434, 23], [382, 418, 23],
# [385, 437, 23],
# [118, 267, 23], [209, 420, 23], [259, 434, 23], [295, 328, 23], [401, 420, 23], [277, 236, 24],
# [394, 436, 24],
# [386, 436, 24], [209, 418, 24], [392, 428, 24], [405, 430, 24], [389, 421, 24], [383, 418, 24],
# [259, 432, 24],
# [411, 428, 24], [392, 426, 25], [397, 438, 25], [383, 420, 25], [404, 430, 25], [386, 437, 25],
# [413, 432, 25],
# [258, 429, 26], [387, 440, 26], [387, 417, 26], [396, 439, 26], [412, 423, 26], [388, 439, 27],
# [397, 424, 28],
# [383, 423, 28], [396, 425, 29], [390, 436, 29], [389, 421, 30]]
img_warped_edge = cv2.cvtColor(img_warped_edge, cv2.COLOR_GRAY2RGB)
img_warped_edge_drawn_top = hough_circles_draw(img_warped_edge, all_1)
plt.imshow(img_warped_edge_drawn_top)
print(all_1)
print(len(all_1))
# 区域2
plt.subplot(222)
plt.title("将圆画在Canny后的斜视图上", fontdict=font)
# list_2 = []
# for i in range(len(all_1)):
# list_i = PointTransformUseM((all_1[i][0], all_1[i][1]), M2)
# list_i[0] = int(list_i[0])
# list_i[1] = int(list_i[1])
# list_i.append(all_1[i][2])
# list_2.append(list_i)
# print(list_2)
# drawn_img_2 = hough_circles_draw(img, list_2)
# drawn_img_2 = cv2.cvtColor(drawn_img_2, cv2.COLOR_BGR2RGB)
# plt.imshow(drawn_img_2)
img_warped_edge_drawn_side = cv2.warpPerspective(img_warped_edge_drawn_top,
M2, (683, 512))
# img_warped_edge_drawn_side = cv2.cvtColor(img_warped_edge_drawn_side, cv2.COLOR_BGR2RGB)
plt.imshow(img_warped_edge_drawn_side)
# 区域3
plt.subplot(223)
plt.title("将圆画在原始图的俯视图上", fontdict=font)
img_warped_drawn_top = hough_circles_draw(img_original, all_1)
img_warped_drawn_top = cv2.cvtColor(img_warped_drawn_top,
cv2.COLOR_BGR2RGB)
plt.imshow(img_warped_drawn_top)
# 区域4
plt.subplot(224)
plt.title("将圆画在原始图的斜视图上", fontdict=font)
img_warped_drawn_side = cv2.warpPerspective(img_warped_drawn_top, M2,
(683, 512))
# img_warped_drawn_side = cv2.cvtColor(img_warped_drawn_side, cv2.COLOR_BGR2RGB)
plt.imshow(img_warped_drawn_side)
print('\033[F\r8) Time elapsed: %.2f s' % (time.time() - start_time))
plt.show()