LABORATORIO 08 - Descripción y Reconocimiento

19/7/2019 8- Descripción y reconocimiento- Guia
LABORATORIO 08
Descripcion y reconocimiento
Escuela de Ingeniería Mecatrónica UNT - Curso Procesamiento de Imagenes Digitales
AGUILAR VALDIVIEZO ANDY HANS

IX CILCO
Docente: Ing. Emerson Maximo Asto Rodriguez
1. Objetivos:
El estudiante conoce diversos descriptores de regiones.
El estudiante aplica diversos descriptores para clasificar figuras.
El estudiante obtiene momentos invariantes de una imagen.
2. Instrucciones
Repase el material teorico y el material extra facilitado por el docente, luego implemente los siguientes
enunciados en Python, ASEGURESE QUE LAS GRAFICAS SEAN LO SUFICIENTEMENTE GRANDES
PARA QUE SE VISUALICEN ADECUADAMENTE, solicite ayuda al docente cuando sea necesario.
2.1. Detección de contornos

Analice el codigo a continuación que sirve para obtener y graficar diferentes contornos, luego aplique el
mismo procedimiento para la imagen "flecha.png"
In [1]:
import numpy as np
import matplotlib.pyplot as plt
from skimage import io
from skimage import measure
localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 1/29

In [3]:
#EJEMPLO
# Contruyendo una imagen
x, y = np.ogrid[-np.pi:np.pi:100j, -np.pi:np.pi:100j]
r = np.sin(np.exp((np.sin(x)**3 + np.cos(y)**2))) #imagen construida r
# buscar contornos con valor mayor a 0.8

contours = measure.find_contours(r, 0.5) #Entrega una lista de contornos en el formato
#[(0,np.array([puntos del contorno])]
# Mostrando imagen
fig, ax = plt.subplots()
ax.imshow(r, cmap='gray')
for n, contour in enumerate(contours):

ax.plot(contour[:, 1], contour[:, 0], linewidth=3)
ax.axis('image')
ax.set_xticks([])
ax.set_yticks([])
plt.show()

In [4]:
print(contours)

[array([[5.35004742, 0. ],
[5.15558015, 1. ],
[5. , 1.26856114],
[4.39810785, 2. ],
[4. , 2.29444112],
[3. , 2.75184936],
[2. , 2.9875167 ],
[1.85757398, 3. ],
[1. , 3.05764552],
[0. , 3.06729478]]), array([[ 0. , 46.4569752 ],
[ 1. , 46.46777372],
[ 2. , 46.54303405],
[ 3. , 46.74625351],
[ 3.64796562, 47. ],
[ 4. , 47.19859561],
[ 4.87645073, 48. ],
[ 5. , 48.22555109],
[ 5.30157125, 49. ],
[ 5.30157125, 50. ],
[ 5. , 50.77444891],
[ 4.87645073, 51. ],
[ 4. , 51.80140439],
[ 3.64796562, 52. ],
[ 3. , 52.25374649],
[ 2. , 52.45696595],
[ 1. , 52.53222628],
[ 0. , 52.5430248 ]]), array([[ 0. , 95.93270522],
[ 1. , 95.94235448],
[ 1.85757398, 96. ],
[ 2. , 96.0124833 ],
[ 3. , 96.24815064],
[ 4. , 96.70555888],
[ 4.39810785, 97. ],
[ 5. , 97.73143886],
[ 5.15558015, 98. ],
[ 5.35004742, 99. ]]), array([[33. , 25.37068293],
[32.94439414, 26. ],
[32.74077445, 27. ],
[32.42867855, 28. ],
[32.01777945, 29. ],
[32. , 29.03489813],
[31.46361135, 30. ],
[31. , 30.72247986],
[30.80066278, 31. ],
[30. , 31.99333977],
[29.9937921 , 32. ],
[29. , 32.96860332],
[28.96099307, 33. ],
[28. , 33.7124201 ],
[27.49417655, 34. ],
[27. , 34.26154226],
[26. , 34.60795154],
[25. , 34.7597504 ],
[24. , 34.70892296],
[23. , 34.45817635],
[22. , 34.02052378],
[21.96678115, 34. ],
[21. , 33.36086533],
[20.56294581, 33. ],
[20. , 32.49708587],
[19.52792025, 32. ],
[19. , 31.38981685],
[18.70258937, 31. ],
[18.02200197, 30. ],
[18. , 29.96211924],
[17.49532718, 29. ],
[17.06216828, 28. ],
[17. , 27.80883824],
[16.75819078, 27. ],
[16.56385092, 26. ],
[16.47942099, 25. ],
[16.50764158, 24. ],
[16.6475913 , 23. ],
[16.8947804 , 22. ],
[17. , 21.70020147],
[17.26653769, 21. ],
[17.74686033, 20. ],
[18. , 19.55788366],
[18.35277167, 19. ],
[19. , 18.09945839],
[19.08123105, 18. ],
[19.9885725 , 17. ],
[20. , 16.98844063],
[21. , 16.13330784],
[21.19499887, 16. ],
[22. , 15.49029252],
[23. , 15.03255231],
[23.12520201, 15. ],
[24. , 14.78907376],
[25. , 14.74027903],
[26. , 14.88602501],
[26.33931457, 15. ],
[27. , 15.23827103],
[28. , 15.78548377],
[28.29800604, 16. ],
[29. , 16.54355857],
[29.48748326, 17. ],
[30. , 17.52269058],
[30.40548901, 18. ],
[31. , 18.77522886],
[31.15434896, 19. ],
[31.75278436, 20. ],
[32. , 20.49036113],
[32.2348803 , 21. ],
[32.59774279, 22. ],
[32.85663913, 23. ],
[33. , 23.97758087],
[33.00303761, 24. ],
[33.03017251, 25. ],
[33. , 25.37068293]]), array([[33. , 75.02241913],
[32.85663913, 76. ],
[32.59774279, 77. ],
[32.2348803 , 78. ],
[32. , 78.50963887],
[31.75278436, 79. ],
[31.15434896, 80. ],
[31. , 80.22477114],
[30.40548901, 81. ],
[30. , 81.47730942],
[29.48748326, 82. ],
[29. , 82.45644143],
[28.29800604, 83. ],
[28. , 83.21451623],
[27. , 83.76172897],
[26.33931457, 84. ],
[26. , 84.11397499],
[25. , 84.25972097],
[24. , 84.21092624],
[23.12520201, 84. ],
[23. , 83.96744769],
[22. , 83.50970748],
[21.19499887, 83. ],
[21. , 82.86669216],
[20. , 82.01155937],
[19.9885725 , 82. ],
[19.08123105, 81. ],
[19. , 80.90054161],
[18.35277167, 80. ],
[18. , 79.44211634],
[17.74686033, 79. ],
[17.26653769, 78. ],
[17. , 77.29979853],
[16.8947804 , 77. ],
[16.6475913 , 76. ],
[16.50764158, 75. ],
[16.47942099, 74. ],
[16.56385092, 73. ],
[16.75819078, 72. ],
[17. , 71.19116176],
[17.06216828, 71. ],
[17.49532718, 70. ],
[18. , 69.03788076],
[18.02200197, 69. ],
[18.70258937, 68. ],
[19. , 67.61018315],
[19.52792025, 67. ],
[20. , 66.50291413],
[20.56294581, 66. ],
[21. , 65.63913467],
[21.96678115, 65. ],
[22. , 64.97947622],
[23. , 64.54182365],
[24. , 64.29107704],
[25. , 64.2402496 ],
[26. , 64.39204846],
[27. , 64.73845774],
[27.49417655, 65. ],
[28. , 65.2875799 ],
[28.96099307, 66. ],
[29. , 66.03139668],
[29.9937921 , 67. ],
[30. , 67.00666023],
[30.80066278, 68. ],
[31. , 68.27752014],
[31.46361135, 69. ],
[32. , 69.96510187],
[32.01777945, 70. ],
[32.42867855, 71. ],
[32.74077445, 72. ],
[32.94439414, 73. ],
[33. , 73.62931707],
[33.03017251, 74. ],
[33.00303761, 75. ],
[33. , 75.02241913]]), array([[44.13439904, 99. ],

[44.35982162, 98. ],
[45. , 97.0653972 ],
[45.06729787, 97. ],
[46. , 96.44446397],
[47. , 96.10721586],
[47.70207797, 96. ],
[48. , 95.96521602],
[49. , 95.93391284],
[50. , 95.9314977 ],
[51. , 95.90026946],
[52. , 95.78435683],
[53. , 95.5340633 ],
[54. , 95.1111483 ],
[54.17372084, 95. ],
[55. , 94.56580918],
[55.80387753, 94. ],
[56. , 93.87990613],
[57. , 93.10149541],
[57.10539588, 93. ],
[58. , 92.23821188],
[58.23708253, 92. ],
[59. , 91.30477741],
[59.29152057, 91. ],
[60. , 90.31562012],
[60.29147896, 90. ],
[61. , 89.28068183],
[61.25172184, 89. ],
[62. , 88.20773632],
[62.1821186 , 88. ],
[63. , 87.10326104],
[63.08914418, 87. ],
[63.97156232, 86. ],
[64. , 85.96769399],
[64.81818982, 85. ],
[65. , 84.7857556 ],
[65.65370454, 84. ],
[66. , 83.57902762],
[66.47852003, 83. ],
[67. , 82.34957573],
[67.29003658, 82. ],
[68. , 81.09670077],
[68.08156811, 81. ],
[68.84019924, 80. ],
[69. , 79.76486915],
[69.57330658, 79. ],
[70. , 78.34148751],
[70.25962122, 78. ],
[70.86404757, 77. ],
[71. , 76.68786727],
[71.3770176 , 76. ],
[71.69072437, 75. ],
[71.75469808, 74. ],
[71.56403005, 73. ],
[71.1330425 , 72. ],
[71. , 71.80107822],
[70.57885881, 71. ],
[70. , 70.14214459],
[69.91736196, 70. ],
[69.21303319, 69. ],
[69. , 68.73773003],
[68.4620915 , 68. ],
[68. , 67.42085348],
[67.68439252, 67. ],
[67. , 66.13862449],
[66.89268597, 66. ],
[66.08046987, 65. ],
[66. , 64.90524293],
[65.23949753, 64. ],
[65. , 63.71723347],
[64.38345475, 63. ],
[64. , 62.55051886],
[63.51276204, 62. ],
[63. , 61.40857206],
[62.62534379, 61. ],
[62. , 60.29614915],
[61.7167344 , 60. ],
[61. , 59.21892828],
[60.7799277 , 59. ],
[60. , 58.18316032],
[59.80483183, 58. ],
[59. , 57.1955469 ],
[58.7769109 , 57. ],
[58. , 56.26369326],
[57.67413286, 56. ],
[57. , 55.39768525],
[56.4603439 , 55. ],
[56. , 54.61392805],
[55.07053671, 54. ],
[55. , 53.94452671],
[54. , 53.37937216],
[53.02128336, 53. ],
[53. , 52.98961616],
[52. , 52.70910906],
[51. , 52.57932765],
[50. , 52.54437617],
[49. , 52.54167332],
[48. , 52.50664397],
[47. , 52.37529585],
[46. , 52.08434155],
[45.83220448, 52. ],
[45. , 51.39363273],
[44.64391475, 51. ],
[44.19060417, 50. ],
[44.19060417, 49. ],
[44.64391475, 48. ],
[45. , 47.60636727],
[45.83220448, 47. ],
[46. , 46.91565845],
[47. , 46.62470415],
[48. , 46.49335603],
[49. , 46.45832668],
[50. , 46.45562383],
[51. , 46.42067235],
[52. , 46.29089094],
[53. , 46.01038384],
[53.02128336, 46. ],
[54. , 45.62062784],
[55. , 45.05547329],
[55.07053671, 45. ],
[56. , 44.38607195],
[56.4603439 , 44. ],
[57. , 43.60231475],
[57.67413286, 43. ],
[58. , 42.73630674],
[58.7769109 , 42. ],
[59. , 41.8044531 ],
[59.80483183, 41. ],
[60. , 40.81683968],
[60.7799277 , 40. ],
[61. , 39.78107172],
[61.7167344 , 39. ],
[62. , 38.70385085],
[62.62534379, 38. ],
[63. , 37.59142794],
[63.51276204, 37. ],
[64. , 36.44948114],
[64.38345475, 36. ],
[65. , 35.28276653],
[65.23949753, 35. ],
[66. , 34.09475707],
[66.08046987, 34. ],
[66.89268597, 33. ],
[67. , 32.86137551],
[67.68439252, 32. ],
[68. , 31.57914652],
[68.4620915 , 31. ],
[69. , 30.26226997],
[69.21303319, 30. ],
[69.91736196, 29. ],
[70. , 28.85785541],
[70.57885881, 28. ],
[71. , 27.19892178],
[71.1330425 , 27. ],
[71.56403005, 26. ],
[71.75469808, 25. ],
[71.69072437, 24. ],
[71.3770176 , 23. ],
[71. , 22.31213273],
[70.86404757, 22. ],
[70.25962122, 21. ],
[70. , 20.65851249],
[69.57330658, 20. ],
[69. , 19.23513085],
[68.84019924, 19. ],
[68.08156811, 18. ],
[68. , 17.90329923],
[67.29003658, 17. ],
[67. , 16.65042427],
[66.47852003, 16. ],
[66. , 15.42097238],
[65.65370454, 15. ],
[65. , 14.2142444 ],
[64.81818982, 14. ],
[64. , 13.03230601],
[63.97156232, 13. ],
[63.08914418, 12. ],
[63. , 11.89673896],
[62.1821186 , 11. ],
[62. , 10.79226368],
[61.25172184, 10. ],
[61. , 9.71931817],
[60.29147896, 9. ],
[60. , 8.68437988],
[59.29152057, 8. ],
[59. , 7.69522259],
[58.23708253, 7. ],
[58. , 6.76178812],
[57.10539588, 6. ],
[57. , 5.89850459],
[56. , 5.12009387],
[55.80387753, 5. ],
[55. , 4.43419082],
[54.17372084, 4. ],
[54. , 3.8888517 ],
[53. , 3.4659367 ],
[52. , 3.21564317],
[51. , 3.09973054],
[50. , 3.0685023 ],
[49. , 3.06608716],
[48. , 3.03478398],
[47.70207797, 3. ],
[47. , 2.89278414],
[46. , 2.55553603],
[45.06729787, 2. ],
[45. , 1.9346028 ],
[44.35982162, 1. ],
[44.13439904, 0. ]]), array([[70.54699662, 0. ],
[70.64017241, 1. ],
[70.91405574, 2. ],
[71. , 2.18795199],
[71.51101978, 3. ],
[72. , 3.57828859],
[72.54644778, 4. ],
[73. , 4.27094536],
[74. , 4.5304272 ],
[75. , 4.44306077],
[76. , 4.01779504],
[76.02317075, 4. ],
[77. , 3.04892399],
[77.03542531, 3. ],
[77.55320252, 2. ],
[77.88210129, 1. ],
[77.99644381, 0. ]]), array([[77.31678423, 52. ],
[77.74275478, 51. ],
[77.96759651, 50. ],
[77.96759651, 49. ],
[77.74275478, 48. ],
[77.31678423, 47. ],
[77. , 46.48995481],
[76.56236013, 46. ],
[76. , 45.51924927],
[75. , 45.0235228 ],
[74.77142266, 45. ],
[74. , 44.94049421],
[73.74443661, 45. ],
[73. , 45.22607216],
[72. , 45.89597677],
[71.89996028, 46. ],
[71.16794153, 47. ],
[71. , 47.32792191],
[70.7551692 , 48. ],
[70.57040118, 49. ],
[70.57040118, 50. ],
[70.7551692 , 51. ],
[71. , 51.67207809],
[71.16794153, 52. ],
[71.89996028, 53. ],
[72. , 53.10402323],
[73. , 53.77392784],
[73.74443661, 54. ],
[74. , 54.05950579],
[74.77142266, 54. ],
[75. , 53.9764772 ],
[76. , 53.48075073],
[76.56236013, 53. ],
[77. , 52.51004519],
[77.31678423, 52. ]]), array([[77.99644381, 99. ],
[77.88210129, 98. ],
[77.55320252, 97. ],
[77.03542531, 96. ],
[77. , 95.95107601],
[76.02317075, 95. ],
[76. , 94.98220496],
[75. , 94.55693923],
[74. , 94.4695728 ],
[73. , 94.72905464],
[72.54644778, 95. ],
[72. , 95.42171141],
[71.51101978, 96. ],
[71. , 96.81204801],
[70.91405574, 97. ],
[70.64017241, 98. ],
[70.54699662, 99. ]]), array([[99. , 3.06729478],
[98. , 3.07693742],
[97. , 3.14378904],
[96. , 3.32127365],
[95. , 3.65400572],
[94.34093569, 4. ],
[94. , 4.14828419],
[93. , 4.76181241],
[92.70664554, 5. ],
[92. , 5.49406352],
[91.41576223, 6. ],
[91. , 6.32165061],
[90.26330812, 7. ],
[90. , 7.22210756],
[89.20137761, 8. ],
[89. , 8.18289136],
[88.20205419, 9. ],
[88. , 9.19545586],
[87.24759631, 10. ],
[87. , 10.25275796],
[86.32593747, 11. ],
[86. , 11.34833231],
[85.42863374, 12. ],
[85. , 12.47617836],
[84.54998666, 13. ],
[84. , 13.63103108],
[83.68672793, 14. ],
[83. , 14.80873492],
[82.83799971, 15. ],
[82.00635077, 16. ],
[82. , 16.00806302],
[81.21020407, 17. ],
[81. , 17.27601067],
[80.42441542, 18. ],
[80. , 18.57421655],
[79.65789822, 19. ],
[79. , 19.91380858],
[78.92888607, 20. ],
[78.24395555, 21. ],
[78. , 21.43927541],
[77.62352232, 22. ],
[77.13933562, 23. ],
[77. , 23.5009915 ],
[76.81522916, 24. ],
[76.7407218 , 25. ],
[76.96288358, 26. ],
[77. , 26.07152126],
[77.35802398, 27. ],
[77.93151192, 28. ],
[78. , 28.08881117],
[78.57510275, 29. ],
[79. , 29.56398352],
[79.28950525, 30. ],
[80. , 30.94964004],
[80.03501078, 31. ],
[80.81106609, 32. ],
[81. , 32.22624033],
[81.61375735, 33. ],
[82. , 33.4671123 ],
[82.43246635, 34. ],
[83. , 34.68530459],
[83.26275336, 35. ],
[84. , 35.88004224],
[84.10360533, 36. ],
[84.96483022, 37. ],
[85. , 37.04093253],
[85.86174188, 38. ],
[86. , 38.15907705],
[86.77982237, 39. ],
[87. , 39.24852937],
[87.72414388, 40. ],
[88. , 40.3033516 ],
[88.70275349, 41. ],
[89. , 41.31647472],
[89.72770619, 42. ],
[90. , 42.27921371],
[90.81722756, 43. ],
[91. , 43.17981726],
[92. , 43.99982625],
[92.00026548, 44. ],
[93. , 44.73666014],
[93.46257111, 45. ],
[94. , 45.36078902],
[95. , 45.83589261],
[95.54306133, 46. ],
[96. , 46.17255425],
[97. , 46.37135141],
[98. , 46.44618356],
[99. , 46.4569752 ]]), array([[99. , 52.5430248 ],
[98. , 52.55381644],
[97. , 52.62864859],
[96. , 52.82744575],
[95.54306133, 53. ],
[95. , 53.16410739],
[94. , 53.63921098],
[93.46257111, 54. ],
[93. , 54.26333986],
[92.00026548, 55. ],
[92. , 55.00017375],
[91. , 55.82018274],
[90.81722756, 56. ],
[90. , 56.72078629],
[89.72770619, 57. ],
[89. , 57.68352528],
[88.70275349, 58. ],
[88. , 58.6966484 ],
[87.72414388, 59. ],
[87. , 59.75147063],
[86.77982237, 60. ],
[86. , 60.84092295],
[85.86174188, 61. ],
[85. , 61.95906747],
[84.96483022, 62. ],
[84.10360533, 63. ],
[84. , 63.11995776],
[83.26275336, 64. ],
[83. , 64.31469541],
[82.43246635, 65. ],
[82. , 65.5328877 ],
[81.61375735, 66. ],
[81. , 66.77375967],
[80.81106609, 67. ],
[80.03501078, 68. ],
[80. , 68.05035996],
[79.28950525, 69. ],
[79. , 69.43601648],
[78.57510275, 70. ],
[78. , 70.91118883],
[77.93151192, 71. ],
[77.35802398, 72. ],
[77. , 72.92847874],
[76.96288358, 73. ],
[76.7407218 , 74. ],
[76.81522916, 75. ],
[77. , 75.4990085 ],
[77.13933562, 76. ],
[77.62352232, 77. ],
[78. , 77.56072459],
[78.24395555, 78. ],
[78.92888607, 79. ],
[79. , 79.08619142],
[79.65789822, 80. ],
[80. , 80.42578345],
[80.42441542, 81. ],
[81. , 81.72398933],
[81.21020407, 82. ],
[82. , 82.99193698],
[82.00635077, 83. ],
[82.83799971, 84. ],
[83. , 84.19126508],
[83.68672793, 85. ],
[84. , 85.36896892],
[84.54998666, 86. ],
[85. , 86.52382164],
[85.42863374, 87. ],
[86. , 87.65166769],
[86.32593747, 88. ],
[87. , 88.74724204],
[87.24759631, 89. ],
[88. , 89.80454414],
[88.20205419, 90. ],
[89. , 90.81710864],
[89.20137761, 91. ],
[90. , 91.77789244],
[90.26330812, 92. ],
[91. , 92.67834939],
[91.41576223, 93. ],
[92. , 93.50593648],
[92.70664554, 94. ],
[93. , 94.23818759],
[94. , 94.85171581],
[94.34093569, 95. ],
[95. , 95.34599428],
[96. , 95.67872635],
[97. , 95.85621096],
[98. , 95.92306258],
[99. , 95.93270522]])]
In [194]:
a=io.imread("flecha.png",as_gray=True)
contornos=measure.find_contours(a,0.3)
print(len(contornos))
ax.imshow(a, cmap='gray')
for n, contornos in enumerate(contornos):

ax.plot(contornos[:, 1], contornos[:, 0], linewidth=3)
plt.show()

2.2. Suavizado de poligonos y aproximacion poligonal

Analize el codigo siguiente, luego aproxime el poligono para la figura "cordiode.tif"

In [6]:
#EJEMPLO Poligonos suavizados y aproximados
from skimage.draw import ellipse

from skimage.measure import find_contours, approximate_polygon,subdivide_polygon
hand = np.array([[1.64516129, 1.16145833],

[1.64516129, 1.59375],
[1.35080645, 1.921875],
[1.375, 2.18229167],
[1.68548387, 1.9375],
[1.60887097, 2.55208333],
[1.68548387, 2.69791667],
[1.76209677, 2.56770833],
[1.83064516, 1.97395833],
[1.89516129, 2.75],
[1.9516129, 2.84895833],
[2.01209677, 2.76041667],
[1.99193548, 1.99479167],
[2.11290323, 2.63020833],
[2.2016129, 2.734375],
[2.25403226, 2.60416667],
[2.14919355, 1.953125],
[2.30645161, 2.36979167],
[2.39112903, 2.36979167],
[2.41532258, 2.1875],
[2.1733871, 1.703125],
[2.07782258, 1.16666667]])
new_hand = hand.copy()
for _ in range(5):
new_hand = subdivide_polygon(new_hand, degree=2, preserve_ends=True) #suavizar la c
urva usando B-splines
# approximate subdivided polygon with Douglas-Peucker algorithm

appr_hand = approximate_polygon(new_hand, tolerance=0.02) #Aproxima poligono según la
distancia de tolerancia
print("numero de coordenadas:", len(hand), len(new_hand), len(appr_hand))
fig, (ax1, ax2) = plt.subplots(ncols=2, figsize=(9, 4))
ax1.plot(hand[:, 0], hand[:, 1])

ax1.plot(new_hand[:, 0], new_hand[:, 1])
ax1.plot(appr_hand[:, 0], appr_hand[:, 1])
# Crear 2 elipses
img = np.zeros((800, 800), 'int32') # lienzo para las imagenes
rr, cc = ellipse(250, 250, 180, 230, img.shape)# 250,250 es el centro de la imagen
#180,230 Semiejes menores y mayores
img[rr, cc] = 1
rr, cc = ellipse(600, 600, 150, 90, img.shape)
img[rr, cc] = 1
ax2.imshow(img,cmap="gray")
for contour in find_contours(img, 0.5): #Encuentra los contornos en la imagen
coords = approximate_polygon(contour, tolerance=6)#Poligonos aproximados con distan

cia 2.5
ax2.plot(coords[:, 1], coords[:, 0], '--r', linewidth=2) #dibuja las coordenadas de
l contorno puntos rojos
coords2 = approximate_polygon(contour, tolerance=35) #Poligonos aproximados con di

stancia 39.5 linea verde
ax2.plot(coords2[:, 1], coords2[:, 0], '-g', linewidth=2)
print("Nnumero de coordenadas:", len(contour), len(coords), len(coords2))
ax2.axis((0, 800, 0, 800)) #dimensiona los ejes del lienzo
plt.show()
numero de coordenadas: 22 704 26

Nnumero de coordenadas: 1637 17 9
Nnumero de coordenadas: 957 17 5

In [8]:
#detectar poligono aproximado de la imagen "cardiode.tif"

card=io.imread("cardiode.tif",as_gray=True)
cont=find_contours(card,0.5)
print(cont)#Verificar forma de cont
coords=approximate_polygon(cont[0],tolerance=15)
#print(coords)#Verificar el tamaño de la lista
ax.imshow(card, cmap='gray')
ax.plot(cont[0][:,1],cont[0][:,0], '-r', linewidth=2)
ax.plot(coords[:,1],coords[:,0], '-b', linewidth=2)
plt.show()
[array([[255.5, 147. ],
[255.5, 146. ],
[255.5, 145. ],
...,
[254.5, 148. ],
[255. , 147.5],
[255.5, 147. ]])]
2.3. Obtención de propiedades de regiones

Analize el codigo siguiente,luego obtenga todas las propiedades para la imagen "estrella.png"
dispominbles dentro de measure.regionprops(img_lab)

In [220]:
#EJEMPlO ALGUNAS PROPIEDADES REGIONPROPS

import math
from skimage.draw import ellipse
from skimage.measure import label, regionprops
from skimage.transform import rotate
#Creando imagen de una elipse rotada 15°

image = np.zeros((600, 600))
rr, cc = ellipse(300, 350, 100, 220,img.shape)
image[rr, cc] = 1
image = rotate(image, angle=15)
label_img = label(image) #imagen etiquetada

regions = regionprops(label_img) #Obtencion de las propiedades
ax.imshow(image, cmap="gray")# mostrando elipse
for props in regions:

y0, x0 = props.centroid #obtencion del centroide
orientation = props.orientation
x1 = x0 + math.cos(orientation) * 0.5 * props.major_axis_length #calculo del punto
a(x,) de la elipse
y1 = y0 - math.sin(orientation) * 0.5 * props.major_axis_length #calculo del punto
a(,y) de la elipse
x2 = x0 - math.sin(orientation) * 0.5 * props.minor_axis_length #calculo del punto
b(x,) de la elipse
y2 = y0 - math.cos(orientation) * 0.5 * props.minor_axis_length #calculo del punto
b(,y) de la elipse
ax.plot((x0, x1), (y0, y1), '-r', linewidth=2.5) # dibujo del semieje mayor
ax.plot((x0, x2), (y0, y2), '-r', linewidth=2.5) # dibujo del semieje menor
ax.plot(x0, y0, '.g', markersize=15) #dibujo del centroide
minr, minc, maxr, maxc = props.bbox # minr y minc son la esquina superior, y maxr
y maxc son la esquina inferior
bx = (minc, maxc, maxc, minc, minc) # ordenando cordenadas filas
by = (minr, minr, maxr, maxr, minr) # ordenando cordenadas columnas filas
ax.plot(bx, by, '-b', linewidth=2.5) # dibujando el rectangulo
ax.axis((0, 600, 600, 0)) #estableciendo la escala de los ejes en pixeles

plt.show()


In [195]:
pent=io.imread("estrella.png",as_gray=True)
lab=measure.label(pent)
propiedades=measure.regionprops(lab)
fig,ax=plt.subplots(1,4,figsize=(16,4))
ax[0].imshow(lab,cmap='gray')
for props in propiedades:

print("El area del poligono es:", props.area)
minr, minc, maxr, maxc = props.bbox # minr y minc son la esquina superior, y maxr
y maxc son la esquina inferior
bx = (minc, maxc, maxc, minc, minc) # ordenando cordenadas filas
by = (minr, minr, maxr, maxr, minr) # ordenando cordenadas columnas filas
ax[0].plot(bx, by, '-b', linewidth=2.5) # dibujando el rectangulo
y0, x0 = props.centroid #obtencion del centroide

print("El centroide es: ", props.centroid)
ax[0].plot(x0, y0, '.g', markersize=15) #dibujo del centroide
print("el area del cuadrilatero azul es: ",props.bbox_area)
ax[1].imshow(props.convex_image,cmap='gray')#convex hull
co=props.coords #coordenadas de los puntos
print("las coordenadas de los pontos son:\n ",co)
excent=props.eccentricity
print("la excenticidad es: ", excent)
print("El diametro del circulo con area equivalente a la region es:",props.equivale
nt_diameter)
print("El numero de euler de la region es: ",props.euler_number)
print("Area de la region/area del bounding box: ", props.extent)
print("El area sin huecos es: ", props.filled_area)
ax[2].imshow(props.filled_image,cmap='gray')#imagen sin huecos
ax[3].imshow(props.image,cmap='gray')
print("tensor de inercia\n ", props.inertia_tensor)
print("valores propios del tensor de inercia: ",props.inertia_tensor_eigvals)
print("Numero de etiqueta: ",props.label)
print("centroide local, respecto al bbox: ",props.local_centroid)
print("Eje mayor de la elipse que tiene el mismo momento central que la region: ",p
rops.major_axis_length)
print("Eje menor de la elipse que tiene el mismo momento central que la region: ",p
rops.minor_axis_length)
print("Los momentos hasta el 3° orden son:\n",props.moments) #Informacion https://e
n.wikipedia.org/wiki/Image_moment
print("Los momentos centrales hasta el 3° orden son:\n",props.moments_central)
print("Los momentos hu:\n",props.moments_hu) # Momentos invariantes a la rotacion,
traslacion y escalamiento
print("Los momentos normalizados:\n",props.moments_normalized)
print("Orientacion de la elipse que tiene el mismo momento central que la region: "
,props.orientation)
print("El perimetro es: ",props.perimeter)
print("Area region/area Convex Hull: ",props.solidity)# solidez de la region
plt.show()

El area del poligono es: 10138

El centroide es: (177.74847109883606, 244.3657526139278)
el area del cuadrilatero azul es: 22010
las coordenadas de los pontos son:
[[101 245]
[102 245]
[103 244]
...
[254 244]
[254 245]
[255 245]]
la excenticidad es: 0.32395692655562325
El diametro del circulo con area equivalente a la region es: 113.613830604
04697
El numero de euler de la region es: -1
Area de la region/area del bounding box: 0.4606088141753748
El area sin huecos es: 11259
tensor de inercia
[[1007.60293956 54.16949407]
[ 54.16949407 1041.34087486]]
valores propios del tensor de inercia: [1081.2072238731917, 967.736590548
7451]
Numero de etiqueta: 1
centroide local, respecto al bbox: (76.74847109883606, 70.3657526139278)
Eje mayor de la elipse que tiene el mismo momento central que la region:
131.52686258696764
Eje menor de la elipse que tiene el mismo momento central que la region:
124.43385973592525
Los momentos hasta el 3° orden son:
[[1.01380000e+04 7.78076000e+05 6.99312220e+07 6.94118152e+09]
[7.13368000e+05 5.42007330e+07 4.83618621e+09 4.78424935e+11]
[6.07537900e+07 4.58386764e+09 4.03682347e+11 3.94106285e+13]
[5.76127242e+09 4.33085998e+11 3.76571673e+13 3.62093865e+15]]
Los momentos centrales hasta el 3° orden son:
[[ 1.01380000e+04 1.63709046e-10 1.02150786e+07 6.08382382e+06]
[-2.91038305e-11 -5.49170331e+05 -2.80891713e+05 -2.27473256e+08]
[ 1.05571138e+07 -1.60728767e+06 7.39363203e+09 2.77051614e+09]
[ 5.67746070e+05 -5.86171827e+08 -1.04254199e+07 -2.40967720e+11]]
Los momentos hu:
[ 2.02105328e-01 1.25274415e-04 1.12914192e-06 1.87889452e-07
8.21328002e-14 -8.76431535e-10 2.72718395e-14]
Los momentos normalizados:
[[ nan nan 9.93887295e-02 5.87889782e-04]
[ nan -5.34321307e-03 -2.71430227e-05 -2.18309927e-04]
[ 1.02716598e-01 -1.55314820e-04 7.09579356e-03 2.64075239e-05]
[ 5.48622254e-05 -5.62559005e-04 -9.93712040e-08 -2.28112835e-05]]
Orientacion de la elipse que tiene el mismo momento central que la region:
0.6344520127784165
El perimetro es: 744.8315651863259
Area region/area Convex Hull: 0.6169293494797055

2.4. Propiedades para imagenes en escala de grises

Analice el siguiente codigo, en el cual se aplica regionprops para la imagen en escala de grises
"flecha.png"

In [195]:
#Propiedades para imagenes en escala de grises

img=io.imread("flecha.png",as_gray=True)
lab=measure.label(img>0.2)
props=measure.regionprops(lab,intensity_image=img,)
plt.subplot(1,2,1)
plt.imshow(img,cmap='gray')
plt.subplot(1,2,2)
plt.imshow(props[0].intensity_image,cmap='gray')
print("La intensidad Maxima es: ",props[0].max_intensity)

print("La intensidad Minima es: ",props[0].min_intensity)
print("La intensidad Media es: ",props[0].mean_intensity)
print("Centroide ponderado: ",props[0].weighted_centroid)
print("Centroide ponderado: ",props[0].weighted_local_centroid)
print("Momentos ponderados:\n ",props[0].weighted_moments)
print("Momentos ponderados centrales:\n ",props[0].weighted_moments_central)
print("Momentos ponderados Hu:\n ",props[0].weighted_moments_hu)
print("Momentos ponderados Normalizados:\n ",props[0].weighted_moments_normalized)

La intensidad Maxima es: 0.8914286274509804

La intensidad Minima es: 0.2862325490196078
La intensidad Media es: 0.35881085217572206
Centroide ponderado: (161.6688686241441, 241.60320552368592)
Centroide ponderado: [50.66886862 84.60320552]
Momentos ponderados:
[[3.11124890e+03 2.63221630e+05 2.77508382e+07 3.37652307e+09]
[1.57643462e+05 1.33149243e+07 1.40222741e+09 1.70511914e+11]
[8.80702959e+06 7.43436358e+08 7.77831106e+10 9.39593892e+12]
[5.29130461e+08 4.46761348e+10 4.63089644e+12 5.53847149e+14]]
Momentos ponderados centrales:
[[ 3.11124890e+03 1.45519152e-11 5.48144450e+06 1.01217658e+08]
[ 2.82227575e-11 -2.22179088e+04 -1.16754268e+05 -6.59694416e+07]
[ 8.19413741e+05 5.84936134e+05 9.66140687e+08 3.13266133e+10]
[-1.49298860e+05 -7.79009592e+06 -2.41947494e+08 -2.66749209e+10]]
Momentos ponderados Hu:
[ 6.50923869e-01 2.31980082e-01 3.39351509e-02 3.55505949e-02
1.23477505e-03 1.71225403e-02 -7.18617319e-06]
Momentos ponderados Normalizados:
[[ nan nan 5.66272549e-01 1.87464653e-01]
[ nan -2.29526941e-03 -2.16239922e-04 -2.19047621e-03]
[ 8.46513192e-02 1.08335692e-03 3.20801289e-02 1.86484096e-02]
[-2.76515577e-04 -2.58665517e-04 -1.44028846e-04 -2.84684693e-04]]
2.5. Propiedades de conjuntos de figuras

Aplique regionprops para la imagen "formas.png", luego etiquete, y flitre las formas segun una
propiedad especifica
Clasifique las formas en cuadrados, rectangulos, elipses, triangulos y rombos

In [10]:
#Propiedades de un conjunto de formas

img=io.imread("formas.png",as_gray=True)
img_bin=img>0.2
img_lab=measure.label(img_bin)
ax[0].imshow(img,cmap='nipy_spectral')
ax[1].imshow(img_bin,cmap='gray')
ax[2].imshow(img_lab==6,cmap='gray')
plt.show()
In [11]:
%matplotlib inline
props=measure.regionprops(img_lab)
ax[0].imshow(img_lab)
ax[0].set_title('Numero de cada figura')
for n,ppd in enumerate(props):

txt=str(n+1)
x,y=ppd.centroid
ax[0].text(y,x,txt,bbox=dict(facecolor='white', alpha=0.5))
ax[1].imshow(img_lab)
ax[1].set_title('Perimetro de cada figura etiquetada')
txt=str(round(ppd.perimeter))
x,y=ppd.centroid
ax[1].text(y-20,x,txt,bbox=dict(facecolor='white', alpha=0.8))

In [15]:
fig,ax =plt.subplots(figsize=(18,5))
lienzo=np.zeros(img.shape)
if (ppd.perimeter**2)/ppd.area<15:
lienzo=lienzo+(img_lab==(n+1))
ax.imshow(lienzo,cmap='gray')
plt.show()
In [ ]:
2.6. Momentos invariantes de Hu

Auna imagen en escala de grises, rotela, escalela, trasladela y luego compruebe la invarianza de los
momentos de Hu

In [287]:
#Moments Hu para imagenes en escala de grises
from skimage import transform
img=io.imread("cañoCristales.jpg",as_gray="True")
img2=transform.rotate(img,30,resize=True)
img3=transform.rescale(img, 0.3)
img4=np.pad(img, ((0,100), (100, 0)), 'constant',constant_values=0)
ax[0,0].imshow(img,cmap='gray')
ax[0,1].imshow(img2,cmap='gray')
muij1=measure.moments_central(img)
muij2=measure.moments_central(img2)
nuij1=measure.moments_normalized(muij1)
print("momentos de HU de la imagen 1", measure.moments_hu(nuij1),"\n")


momentos de HU de la imagen 1 [ 3.33972425e-01 2.60456211e-02 2.87451134

e-04 1.69702762e-04
-3.74759122e-08 1.91691088e-05 -6.39214655e-10]

e-04 1.69740447e-04
-3.74835731e-08 1.91745559e-05 -6.50268044e-10]

e-04 1.68843998e-04
-3.71214599e-08 1.91903640e-05 -5.05859680e-10]

e-04 1.69702762e-04
-3.74759122e-08 1.91691088e-05 -6.39214655e-10]
3. Homework
Complete la guía
RECUERDE LA PRESENTACION ES A TRAVÉS DEL GOOGLE CLASROOM
El nombre del archivo tiene el siguiente formato: (1°apellido)-(1° o 1° Nombre) por ejemplo
Asto-Emerson
El formato tiene que estar importado a PDF
No enviar codigos u otro archivo
La fecha de presentacion cuenta en su calificacion
Los trabajos iguales comparten nota (si hay 2 trabajos iguales y la nota del trabajo es 16, ambos
estudiantes reciben calificacion 8)
In [ ]:

LABORATORIO 08 - Descripción y Reconocimiento

Cargado por

Copyright:

Formatos disponibles

LABORATORIO 08 - Descripción y Reconocimiento

Cargado por

Información del documento

Descripción original:

Título original

Derechos de autor

Formatos disponibles

Compartir este documento

Compartir o incrustar documentos

Opciones para compartir

¿Le pareció útil este documento?

¿Este contenido es inapropiado?

Copyright:

Formatos disponibles

LABORATORIO 08 - Descripción y Reconocimiento

Cargado por

Copyright:

Formatos disponibles

19/7/2019 8- Descripción y reconocimiento- Guia

Escuela de Ingeniería Mecatrónica UNT - Curso Procesamiento de Imagenes Digitales

AGUILAR VALDIVIEZO ANDY HANS

Docente: Ing. Emerson Maximo Asto Rodriguez

2.1. Detección de contornos

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 1/29

# buscar contornos con valor mayor a 0.8

for n, contour in enumerate(contours):

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 2/29

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 3/29

[33. , 75.02241913]]), array([[44.13439904, 99. ],

for n, contornos in enumerate(contornos):

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 14/29

2.2. Suavizado de poligonos y aproximacion poligonal

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 15/29

#EJEMPLO Poligonos suavizados y aproximados

from skimage.draw import ellipse

hand = np.array([[1.64516129, 1.16145833],

# approximate subdivided polygon with Douglas-Peucker algorithm

print("numero de coordenadas:", len(hand), len(new_hand), len(appr_hand))

fig, (ax1, ax2) = plt.subplots(ncols=2, figsize=(9, 4))

ax1.plot(hand[:, 0], hand[:, 1])

coords = approximate_polygon(contour, tolerance=6)#Poligonos aproximados con distan

coords2 = approximate_polygon(contour, tolerance=35) #Poligonos aproximados con di

print("Nnumero de coordenadas:", len(contour), len(coords), len(coords2))

ax2.axis((0, 800, 0, 800)) #dimensiona los ejes del lienzo

numero de coordenadas: 22 704 26

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 17/29

#detectar poligono aproximado de la imagen "cardiode.tif"

print(cont)#Verificar forma de cont

2.3. Obtención de propiedades de regiones

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 18/29

#EJEMPlO ALGUNAS PROPIEDADES REGIONPROPS

#Creando imagen de una elipse rotada 15°

label_img = label(image) #imagen etiquetada

for props in regions:

ax.axis((0, 600, 600, 0)) #estableciendo la escala de los ejes en pixeles

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 19/29

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 20/29

for props in propiedades:

y0, x0 = props.centroid #obtencion del centroide

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 21/29

El area del poligono es: 10138

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 22/29

2.4. Propiedades para imagenes en escala de grises

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 23/29

#Propiedades para imagenes en escala de grises

print("La intensidad Maxima es: ",props[0].max_intensity)

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 24/29

La intensidad Maxima es: 0.8914286274509804

2.5. Propiedades de conjuntos de figuras

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 25/29

#Propiedades de un conjunto de formas

for n,ppd in enumerate(props):

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 26/29

for n,ppd in enumerate(props):

2.6. Momentos invariantes de Hu

localhost:8888/nbconvert/html/Desktop/PROCESAMIENTO DE IMAGENES DIGITALES/CLASE 08_DESCRIPCION Y RECONOCIMIENTO/8- … 27/29