File size: 4,117 Bytes
436b829 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 | import os
import cv2
import numpy as np
from ppd.utils.logger import Log
import torch
import torch.nn.functional as F
from omegaconf.listconfig import ListConfig
EPS = 1e-4
class PrepareForNet(object):
"""Prepare sample for usage as network input.
"""
def __init__(self):
pass
def __str__(self):
return "PrepareForNet"
def __repr__(self):
return "PrepareForNet"
def __call__(self, sample):
image = np.transpose(sample["image"], (2, 0, 1))
sample["image"] = np.ascontiguousarray(image).astype(np.float32)
if "mask" in sample:
sample["mask"] = sample["mask"].astype(np.uint8)
sample["mask"] = np.ascontiguousarray(sample["mask"])[None]
if "depth" in sample:
depth = sample["depth"].astype(np.float32)
sample["depth"] = np.ascontiguousarray(depth)[None]
return sample
def cv2_resize(image, size, interpolation=cv2.INTER_LINEAR):
return cv2.resize(image, size, interpolation=interpolation)[None]
class Resize(object):
"""Resize sample to given size (width, height).
"""
def __init__(
self,
width=None,
height=None,
# image_interpolation_method=cv2.INTER_AREA,
image_interpolation_method = cv2.INTER_LINEAR,
):
self.width = width
self.height = height
self.__image_interpolation_method = image_interpolation_method
def __call__(self, sample):
width, height = self.width, self.height
if width == sample['image'].shape[1] and height == sample['image'].shape[0]:
return sample
Log.debug(
'Resize: {} -> {}'.format(sample["image"].shape, (height, width)))
# resize sample
ori_height, ori_width = sample['image'].shape[:2]
sample["image"] = cv2.resize(
sample["image"],
(width, height),
interpolation=self.__image_interpolation_method,
)
if "depth" in sample:
sample["depth"] = cv2.resize(
sample["depth"],
(width, height),
interpolation=cv2.INTER_NEAREST)
if "mask" in sample:
sample["mask"] = cv2.resize(
sample["mask"].astype(np.float32),
(width, height),
interpolation=cv2.INTER_NEAREST)
return sample
class Resize_4K_Crop(object):
"""Resize sample to given size (width, height).
"""
def __init__(
self,
width=None,
height=None,
crop_type='random',
image_interpolation_method=cv2.INTER_AREA,
):
self.width = width
self.height = height
self.crop_type = crop_type
self.__image_interpolation_method = image_interpolation_method
def __call__(self, sample):
width, height = 1920, 1024
sample["image"] = cv2.resize(
sample["image"],
(width, height),
interpolation=self.__image_interpolation_method,
)
# crop sample
crop_h = self.height
crop_w = self.width
if self.crop_type == 'random':
# random crop
top = np.random.randint(0, height - crop_h + 1)
left = np.random.randint(0, width - crop_w + 1)
else:
# center crop
top = (height - crop_h) // 2
left = (width - crop_w) // 2
sample["image"] = sample["image"][top:top+crop_h, left:left+crop_w]
if "depth" in sample:
sample["depth"] = cv2.resize(
sample["depth"], (width, height),
interpolation=cv2.INTER_NEAREST
)
# crop sample
sample["depth"] = sample["depth"][top:top+crop_h, left:left+crop_w]
if "mask" in sample:
sample["mask"] = cv2.resize(
sample["mask"].astype(np.float32),
(width, height),
interpolation=cv2.INTER_NEAREST,
)
sample["mask"] = sample["mask"][top:top+crop_h, left:left+crop_w]
return sample |