|
import numpy as np |
|
import json |
|
import pdb |
|
from matplotlib import pyplot as plt |
|
import os |
|
|
|
|
|
from scipy.ndimage.morphology import binary_dilation, binary_erosion, binary_hit_or_miss |
|
import random |
|
|
|
from ListSelEm import * |
|
from Utils import Process, Change_Colour |
|
|
|
|
|
""" |
|
Example program for Category A Simple |
|
""" |
|
|
|
print("--------------------------") |
|
print("------ CAT A SIMPLE ------") |
|
print("--------------------------") |
|
|
|
|
|
def _perform_CatA_Simple(img, op, se): |
|
list_se = ['SE1', 'SE2', 'SE3', 'SE4', 'SE5', 'SE6', 'SE7', 'SE8'] |
|
list_se_idx = list_se.index(se) |
|
if op == 'Dilation': |
|
return binary_dilation(img, list_se_3x3[list_se_idx]) |
|
elif op == 'Erosion': |
|
return binary_erosion(img, list_se_3x3[list_se_idx]) |
|
|
|
|
|
idx_select = np.random.randint(100) |
|
|
|
|
|
with open("./Dataset/CatA_Simple/Task{:03d}.json".format(idx_select), 'r') as f: |
|
data = json.load(f) |
|
|
|
|
|
with open("./Dataset/CatA_Simple/Task{:03d}_soln.txt".format(idx_select), 'r') as f: |
|
list_ops = f.readlines() |
|
list_ops = [x.split() for x in list_ops] |
|
|
|
for d in data: |
|
img = np.array(d['input'], dtype=np.int32) |
|
for op, se in list_ops: |
|
img = _perform_CatA_Simple(img, op, se) |
|
img = img*1 |
|
|
|
out = np.array(d['output'], dtype=np.int32) |
|
check_same = np.all(img == out) |
|
if check_same: |
|
print("Program works!!") |
|
else: |
|
print("Something went wrong!!") |
|
|
|
|
|
""" |
|
Example program for Category A Hard |
|
""" |
|
|
|
print("--------------------------") |
|
print("------- CAT A HARD -------") |
|
print("--------------------------") |
|
|
|
|
|
def _perform_CatA_Hard(img, band, op, se): |
|
if band is not None: |
|
list_se = ['SE1', 'SE2', 'SE3', 'SE4', 'SE5', 'SE6', 'SE7', 'SE8'] |
|
list_se_idx = list_se.index(se) |
|
if op == 'Dilation': |
|
return binary_dilation(img, list_se_3x3[list_se_idx]) |
|
elif op == 'Erosion': |
|
return binary_erosion(img, list_se_3x3[list_se_idx]) |
|
|
|
else: |
|
return Change_Colour(img, np.array(se, dtype=np.int32)) |
|
|
|
|
|
idx_select = np.random.randint(100) |
|
|
|
|
|
with open("./Dataset/CatA_Hard/Task{:03d}.json".format(idx_select), 'r') as f: |
|
data = json.load(f) |
|
|
|
|
|
with open("./Dataset/CatA_Hard/Task{:03d}_soln.json".format(idx_select), 'r') as f: |
|
list_ops = json.load(f) |
|
|
|
for d in data: |
|
img = np.array(d['input'], dtype=np.int32) |
|
img = Process(img, num_colors=3) |
|
for band, op, se in list_ops: |
|
if band is not None: |
|
img[:, :, band-1] = _perform_CatA_Hard(img[:, :, band-1], band, op, se) |
|
else: |
|
img = _perform_CatA_Hard(img, band, op, se) |
|
img = img*1 |
|
|
|
out = np.array(d['output'], dtype=np.int32) |
|
check_same = np.all(img == out) |
|
if check_same: |
|
print("Program works!!") |
|
else: |
|
print("Something went wrong!!") |
|
|
|
""" |
|
Example program for Category B Iteration |
|
""" |
|
|
|
print("---------------------------") |
|
print("------ CAT B ITERATE ------") |
|
print("---------------------------") |
|
|
|
|
|
def _perform_CatB_Iteration(img, n_iterate, op, se): |
|
for _ in range(n_iterate): |
|
list_se = ['SE1', 'SE2', 'SE3', 'SE4', 'SE5', 'SE6', 'SE7', 'SE8'] |
|
list_se_idx = list_se.index(se) |
|
if op == 'Dilation': |
|
img = binary_dilation(img, list_se_3x3[list_se_idx]) |
|
elif op == 'Erosion': |
|
img = binary_erosion(img, list_se_3x3[list_se_idx]) |
|
return img |
|
|
|
|
|
|
|
idx_select = 0 |
|
with open("./Dataset/CatB_Iteration/Task{:03d}.json".format(idx_select), 'r') as f: |
|
data = json.load(f) |
|
|
|
|
|
with open("./Dataset/CatB_Iteration/Task{:03d}_soln.json".format(idx_select), 'r') as f: |
|
list_ops = json.load(f) |
|
|
|
for d in data: |
|
img = np.array(d['input'], dtype=np.int32) |
|
for subtask, n_iterate, op, se in list_ops: |
|
if d['subtask'] == subtask: |
|
img = _perform_CatB_Iteration(img, n_iterate, op, se) |
|
img = img*1 |
|
|
|
out = np.array(d['output'], dtype=np.int32) |
|
check_same = np.all(img == out) |
|
if check_same: |
|
print("Program works!!") |
|
else: |
|
print("Something went wrong!!") |
|
|
|
""" |
|
Example program for Category B Sequence |
|
""" |
|
|
|
print("---------------------------") |
|
print("------ CAT B Sequence ------") |
|
print("---------------------------") |
|
|
|
|
|
def _perform_CatB_sequence(img, op, se): |
|
list_se = ['SE1', 'SE2', 'SE3', 'SE4', 'SE5', 'SE6', 'SE7', 'SE8'] |
|
list_se_idx = list_se.index(se) |
|
if op == 'Dilation': |
|
img = binary_dilation(img, list_se_3x3[list_se_idx]) |
|
elif op == 'Erosion': |
|
img = binary_erosion(img, list_se_3x3[list_se_idx]) |
|
|
|
return img |
|
|
|
|
|
|
|
idx_select = 0 |
|
with open("./Dataset/CatB_Sequence/Task{:03d}.json".format(idx_select), 'r') as f: |
|
data = json.load(f) |
|
|
|
|
|
with open("./Dataset/CatB_Sequence/Task{:03d}_soln.json".format(idx_select), 'r') as f: |
|
list_ops = json.load(f) |
|
|
|
for d in data: |
|
img = np.array(d['input'], dtype=np.int32) |
|
for subtask, op, se in list_ops: |
|
if d['subtask'] == subtask: |
|
img = _perform_CatB_sequence(img, op, se) |
|
img = img*1 |
|
|
|
out = np.array(d['output'], dtype=np.int32) |
|
check_same = np.all(img == out) |
|
if check_same: |
|
print("Program works!!") |
|
else: |
|
print("Something went wrong!!") |
|
|
|
|
|
""" |
|
Example program for Category B Selection |
|
""" |
|
|
|
print("-----------------------------") |
|
print("------ CAT B Selection ------") |
|
print("-----------------------------") |
|
|
|
|
|
def _perform_CatB_selection(img, op, se): |
|
list_se = ['SE1', 'SE2', 'SE3', 'SE4', 'SE5', 'SE6', 'SE7', 'SE8'] |
|
if op == 'Dilation': |
|
list_se_idx = list_se.index(se) |
|
img = binary_dilation(img, list_se_3x3[list_se_idx]) |
|
elif op == 'Erosion': |
|
list_se_idx = list_se.index(se) |
|
img = binary_erosion(img, list_se_3x3[list_se_idx]) |
|
elif op == 'Hit-Or-Miss': |
|
list_se_idx = list_se.index(se) |
|
tmp_img = binary_hit_or_miss(img, list_se_3x3[list_se_idx]) |
|
img[tmp_img] = 2 |
|
img = Process(img, num_colors=2) |
|
elif op == 'change_color': |
|
img = Change_Colour(img, np.array(se, dtype=np.int32)) |
|
return img |
|
|
|
|
|
|
|
with open("./Dataset/CatB_Selection/Task{:03d}.json".format(idx_select), 'r') as f: |
|
data = json.load(f) |
|
|
|
|
|
with open("./Dataset/CatB_Selection/Task{:03d}_soln.json".format(idx_select), 'r') as f: |
|
list_ops = json.load(f) |
|
|
|
for d in data: |
|
img = np.array(d['input'], dtype=np.int32) |
|
for band, op, se in list_ops: |
|
if band is not None: |
|
img[:, :, band-1] = _perform_CatB_selection(img[:, :, band-1], op, se) |
|
else: |
|
img = _perform_CatB_selection(img, op, se) |
|
img = img*1 |
|
out = np.array(d['output'], dtype=np.int32) |
|
check_same = np.all(img == out) |
|
if check_same: |
|
print("Program works!!") |
|
else: |
|
print("Something went wrong!!") |
|
|
|
|
|
""" |
|
Example program for Category B Hard |
|
""" |
|
|
|
print("-----------------------------") |
|
print("------ CAT B Hard ------") |
|
print("-----------------------------") |
|
|
|
|
|
def _perform_CatB_hard(img, k_iterate, op, se): |
|
for _ in range(k_iterate): |
|
list_se = ['SE1', 'SE2', 'SE3', 'SE4', 'SE5', 'SE6', 'SE7', 'SE8'] |
|
if op == 'Dilation': |
|
list_se_idx = list_se.index(se) |
|
img = binary_dilation(img, list_se_3x3[list_se_idx]) |
|
elif op == 'Erosion': |
|
list_se_idx = list_se.index(se) |
|
img = binary_erosion(img, list_se_3x3[list_se_idx]) |
|
elif op == 'Hit-Or-Miss': |
|
list_se_idx = list_se.index(se) |
|
tmp_img = binary_hit_or_miss(img, list_se_3x3[list_se_idx]) |
|
img[tmp_img] = 2 |
|
img = Process(img, num_colors=2) |
|
elif op == 'change_color': |
|
img = Change_Colour(img, np.array(se, dtype=np.int32)) |
|
return img |
|
|
|
|
|
|
|
with open("./Dataset/CatB_Hard/Task{:03d}.json".format(idx_select), 'r') as f: |
|
data = json.load(f) |
|
|
|
|
|
with open("./Dataset/CatB_Hard/Task{:03d}_soln.json".format(idx_select), 'r') as f: |
|
list_ops = json.load(f) |
|
|
|
for d in data: |
|
img = np.array(d['input'], dtype=np.int32) |
|
for band, k_iterate, op, se in list_ops: |
|
if band is not None: |
|
img[:, :, band-1] = _perform_CatB_hard(img[:, :, band-1], k_iterate, op, se) |
|
else: |
|
img = _perform_CatB_hard(img, k_iterate, op, se) |
|
img = img*1 |
|
out = np.array(d['output'], dtype=np.int32) |
|
check_same = np.all(img == out) |
|
if check_same: |
|
print("Program works!!") |
|
else: |
|
print("Something went wrong!!") |
|
|