diff --git a/vviewer/Image.py b/vviewer/Image.py
index c80dfdf6cc3dce2f136d06608e938c7dda6acbc0..94cdc90ee38802b1353956ab4c29eeb5148b45f6 100644
--- a/vviewer/Image.py
+++ b/vviewer/Image.py
@@ -17,10 +17,10 @@ import ImageDialog
from resample import resample_image
from quaternions import fillpositive, quat2mat, mat2quat
-try:
- from pyvista import pyvista
-except ImportError:
- pass
+# try:
+# from pyvista import pyvista
+# except ImportError:
+# pass
class Image(object):
diff --git a/vviewer/Image3D.py b/vviewer/Image3D.py
index 234ad7c9f0ac5d876d8996079b2d46de7fa8817e..773e9e74cf910e66e803fd5d6f21bdc22d306159 100644
--- a/vviewer/Image3D.py
+++ b/vviewer/Image3D.py
@@ -2,10 +2,10 @@ import pyqtgraph as pg
from Image import Image
import ColorMapWidget
-try:
- from pyvista import pyvista
-except ImportError:
- pass
+# try:
+# from pyvista import pyvista
+# except ImportError:
+# pass
class Image3D(Image):
diff --git a/vviewer/Image4D.py b/vviewer/Image4D.py
index 33b5996090bd5a945fe2dc863b80255e5d4ee201..1f93002f200f93263d1d2f3f5093a1785c8dcff7 100644
--- a/vviewer/Image4D.py
+++ b/vviewer/Image4D.py
@@ -16,10 +16,10 @@ from Image import Image
from resample import resample_image
from TimePlot import TimePlot
from testInputs import testFloat, testInteger
-try:
- from pyvista import pyvista
-except ImportError:
- pass
+# try:
+# from pyvista import pyvista
+# except ImportError:
+# pass
class Image4D(Image):
diff --git a/vviewer/VistaLoad.py b/vviewer/VistaLoad.py
new file mode 100644
index 0000000000000000000000000000000000000000..e6c1370212a7cc348e58b4cbb082d6e4a77cc726
--- /dev/null
+++ b/vviewer/VistaLoad.py
@@ -0,0 +1,258 @@
+# -*- coding: utf-8 -*-
+"""
+vista loader, native python implementation
+"""
+
+import scipy.io as sio
+import numpy as np
+import subprocess
+import nibabel as nib
+import re
+
+
+#%%writer
+# dtype = np.int16
+# # data_orig = 13*np.ones((6,6,6), dtype=dtype)
+# xdim = 2
+# ydim = 5
+# zdim = 4
+# # tdim = 11
+
+# data_1D = np.arange(xdim*ydim*zdim, dtype=np.int32)
+# data_orig = np.reshape(data_1D, (xdim, ydim, zdim))
+
+# # data_1D = np.arange(xdim*ydim*zdim*tdim, dtype=dtype)
+# # data_orig = np.reshape(data_1D, (xdim, ydim, zdim, tdim))
+
+
+# nii_orig = nib.Nifti1Image(data_orig, np.eye(4))
+# nii_orig.to_filename("/home/morty/tmp/test.nii")
+# subprocess.call("vnifti -in /home/morty/tmp/test.nii -out /home/morty/tmp/test.v", shell=True)
+
+# #%%BINARY writer
+# dtype = np.int16
+# # data_orig = 13*np.ones((6,6,6), dtype=dtype)
+# xdim = 5
+# ydim = 5
+# zdim = 5
+# # tdim = 11
+
+# data_1D = np.ones(xdim*ydim*zdim, dtype=np.int32)
+# data_1D = np.random.rand(xdim*ydim*zdim)
+# data_1D[data_1D>0.5] = 1
+# data_1D[data_1D<=0.5] = 0
+# # data_1D = np.zeros(xdim*ydim*zdim)
+# # data_1D[4] = 1
+
+
+
+# data_orig = np.reshape(data_1D, (xdim, ydim, zdim))
+
+# nii_orig = nib.Nifti1Image(data_orig, np.eye(4))
+# nii_orig.to_filename("/home/morty/tmp/test.nii")
+# subprocess.call("vnifti -in /home/morty/tmp/test.nii -out /home/morty/tmp/testa.v", shell=True)
+# subprocess.call("vbinarize -in /home/morty/tmp/testa.v -out /home/morty/tmp/test.v -min 0.5", shell=True)
+
+
+#%% AUX
+def get_property_str(header_img, str_property, repn_property="int"):
+ idx_begin = header_img.find(str_property)
+ if idx_begin == -1:
+ return -1
+ idx_val = idx_begin + len(str_property) + 2
+ value = header_img[idx_val:].split('\n')[0]
+ if repn_property == "int":
+ value = int(value)
+ elif repn_property == "str":
+ pass
+ elif repn_property == "float":
+ value = float(value)
+ elif repn_property == "list":
+ value = value[1:-1].split(" ")
+ value = [float(l) for l in value]
+ return value
+
+def get_subheader(header, section_demark):
+ idx_start = header.find(section_demark)
+ if idx_start == -1:
+ return
+ idx_stop = -1
+ for j in range(idx_start, idx_start+5000):
+ if header[j] == "}":
+ idx_stop = j+1
+ break
+ return header[idx_start:idx_stop]
+
+
+#%% reader
+fp_input = "/home/morty/tmp/test.v"
+
+def load_vista(fp_input):
+
+ with open(fp_input, 'rb') as f:
+ raw=f.read()
+
+ #find the ^L breaker, determining where the header part stops
+ for i in range(len(raw)):
+ if raw[i:i+2]==b'\x0c\n':
+ last_idx_header = i+2
+ break
+
+ #now parse the header for the first time to load the data.
+ header = raw[0:last_idx_header-2].decode("utf-8")
+
+ #find all proper images and parse each of them into dict. store all dicts in list_images
+ list_imagedict = []
+ idx_images = [m.start() for m in re.finditer('image: image {', header)]
+
+ for i in range(len(idx_images)):
+ dict_image = {}
+ dict_image['idx_header_begin'] = idx_images[i]
+ #find end of image definition in header
+ for j in range(idx_images[i], idx_images[i]+5000):
+ if header[j] == "}":
+ dict_image['idx_header_end'] = j+1
+ break
+
+ header_img = header[dict_image['idx_header_begin']:dict_image['idx_header_end']]
+ dict_image["repn"] = get_property_str(header_img, "repn", "str")
+ dict_image["offset"] = get_property_str(header_img, "data", "int") + last_idx_header
+ dict_image["length"] = get_property_str(header_img, "length", "int")
+
+ dict_image["nbands"] = get_property_str(header_img, "nbands", "int")
+ if dict_image["nbands"] == -1:
+ dict_image["nbands"] = 1
+
+
+ dict_image["nrows"] = get_property_str(header_img, "nrows", "int")
+ dict_image["ncolumns"] = get_property_str(header_img, "ncolumns", "int")
+ tr = get_property_str(header_img, "repetition_time", "float")
+
+ if tr > 0:
+ dict_image["repetition_time"] = tr
+
+ #fix dtype
+ if dict_image["repn"] == "int":
+ dict_image["dtype"] = np.int32
+ dict_image["length"] = int(dict_image["length"]/4)
+ elif dict_image["repn"] == "long":
+ dict_image["dtype"] = np.int64
+ dict_image["length"] = int(dict_image["length"]/8)
+ elif dict_image["repn"] == "float":
+ dict_image["dtype"] = np.float32
+ dict_image["length"] = int(dict_image["length"]/4)
+ elif dict_image["repn"] == "double":
+ dict_image["dtype"] = np.float64
+ dict_image["length"] = int(dict_image["length"]/8)
+ elif dict_image["repn"] == "short":
+ dict_image["dtype"] = np.int16
+ dict_image["length"] = int(dict_image["length"]/2)
+ elif dict_image["repn"] == "bit":
+ dict_image["dtype"] = np.int16
+ dict_image["length"] = int(dict_image["length"]/1)
+ elif dict_image["repn"] == "ubyte":
+ dict_image["dtype"] = np.uint8
+ dict_image["length"] = int(dict_image["length"]/1)
+
+ else:
+ raise ValueError("Read error: data representation '{}' unknown. please contact support.".format(dict_image["repn"]))
+
+ list_imagedict.append(dict_image)
+
+
+ #decide: was the image 3D or 4D (time series)
+ if len(list_imagedict) == 1: #<=3D case
+ dict_image = list_imagedict[0]
+ xdim = dict_image["ncolumns"]
+ ydim = dict_image["nrows"]
+ zdim = dict_image["nbands"]
+ tdim = 1
+
+ if dict_image["repn"] != "bit": #default case
+ img1D = np.frombuffer(raw, dtype=dict_image["dtype"], count=dict_image["length"], offset=dict_image["offset"]).byteswap()
+ else: #bit representation (masks etc)
+ img1D_byterepn = np.frombuffer(raw, dtype=np.uint8, count=dict_image["length"], offset=dict_image["offset"])#.byteswap()
+ img1D_bit_graced = np.array([])
+ for l in img1D_byterepn:
+ bx = bin(l)[2:]
+ bx = (8-len(bx))*"0"+bx
+ for j in range(8):
+ # b = bx[7-j]
+ b = bx[j]
+ img1D_bit_graced = np.append(img1D_bit_graced, int(b))
+ img1D = img1D_bit_graced[0:xdim*ydim*zdim]
+
+ img3D = np.transpose(np.reshape(img1D, (zdim,ydim,xdim)), (2,1,0))
+ data = img3D
+ dim = "4D"
+ else:
+ list_images = []
+ for i in range(len(idx_images)):
+ dict_image = list_imagedict[i]
+ list_images.append(np.frombuffer(raw, dtype=dict_image["dtype"], count=dict_image["length"], offset=dict_image["offset"]).byteswap())
+ if dict_image["repn"] == "bit":
+ raise ValueError("bit reprensetation not allowed for 4D images!")
+ #concatenate, form a long 1D vector
+ xdim = dict_image["ncolumns"]
+ ydim = dict_image["nrows"]
+ tdim = dict_image["nbands"]
+ zdim = len(list_images)
+ img1D = np.zeros(xdim*ydim*zdim*tdim, dtype=dict_image["dtype"])
+ for i in range(len(idx_images)):
+ img1D[i*xdim*ydim*tdim:(i+1)*xdim*ydim*tdim] = list_images[i]
+ img4D = np.transpose(np.reshape(img1D, (zdim,tdim,ydim,xdim)), (3,2,0,1))
+ data = img4D
+ dim = "3D"
+
+
+ #%% re-parse the header to get the complete header information
+ mm = get_property_str(header_img, "voxel", "list")
+
+ #get s-form code
+ sform_code = get_property_str(header, "sform_code", "int")
+ qform_code = get_property_str(header, "qform_code", "int")
+
+ header_sform = get_subheader(header, "sform: image")
+ offset_sform = get_property_str(header_sform, "data", "int") + last_idx_header
+ length_sform = int(get_property_str(header_sform, "length", "int")/4)
+ sform1D = np.frombuffer(raw, dtype=np.float32, count=length_sform, offset=offset_sform).byteswap()
+ sform2D = np.reshape(sform1D, (4,4))
+
+
+ header_dim = get_subheader(header, "dim: bundle")
+ offset_dim = get_property_str(header_dim, "data", "int") + last_idx_header
+ length_dim = int(get_property_str(header_dim, "length", "int")/4)
+ dim1D = np.frombuffer(raw, dtype=np.float32, count=length_dim, offset=offset_dim)
+
+
+ header_pixdim = get_subheader(header, "pixdim: bundle")
+ offset_pixdim= get_property_str(header_dim, "data", "int") + last_idx_header
+ length_pixdim = int(get_property_str(header_dim, "length", "int")/4)
+ pixdim1D = np.frombuffer(raw, dtype=np.float32, count=length_dim, offset=offset_dim)
+
+
+ header_qform = get_subheader(header, "qform: bundle")
+ offset_qform = get_property_str(header_qform, "data", "int") + last_idx_header
+ length_qform = int(get_property_str(header_qform, "length", "int")/4)
+ qformdim1D = np.frombuffer(raw, dtype=np.float32, count=length_qform, offset=offset_qform)
+ # quatern_b quatern_c quatern_d qoffset_x qoffset_y qoffset_z
+
+
+ #%% convert to nii object... fill in data and header
+ nii_loaded = nib.Nifti1Image(data, affine=np.eye(4))
+ nii_loaded.header.set_zooms(mm)
+ nii_loaded.header['pixdim'][4] = tr
+ nii_loaded.set_sform(sform2D)
+ nii_loaded.header['qform_code'] = qform_code
+ nii_loaded.header['qform_code'] = sform_code
+ nii_loaded.header['quatern_b'] = qformdim1D[0]
+ nii_loaded.header['quatern_c'] = qformdim1D[1]
+ nii_loaded.header['quatern_d'] = qformdim1D[2]
+ nii_loaded.header['qoffset_x'] = qformdim1D[3]
+ nii_loaded.header['qoffset_y'] = qformdim1D[4]
+ nii_loaded.header['qoffset_z'] = qformdim1D[5]
+ nii_loaded.header.set_xyzt_units(xyz="mm", t="msec")
+
+ return nii_loaded
+
+# nii_loaded.to_filename("/home/morty/tmp/out.nii")
diff --git a/vviewer/config.ini b/vviewer/config.ini
index a755b53227378f63e50a7e9d6ef4ab085ab45183..69dab5feeafaf6c56f5d807040bffaa52d7abaea 100644
--- a/vviewer/config.ini
+++ b/vviewer/config.ini
@@ -1,5 +1,5 @@
[view]
-voxel_coord = False
+voxel_coord = True
link_mode = 0
width = 1272
height = 424
@@ -21,5 +21,5 @@ interpolation = 0
os_ratio = 1.0
[search]
-search_radius = 666
+search_radius = 5
diff --git a/vviewer/loadImage.py b/vviewer/loadImage.py
index 18b159cebba80be6ffbe52157dbc6f74efca94ae..0d3eaec6c58228ca3804e836391fd6c8af0000a0 100644
--- a/vviewer/loadImage.py
+++ b/vviewer/loadImage.py
@@ -21,11 +21,12 @@ import ColorMapWidget
import ImageDialog
import Image3D
import Image4D
+from VistaLoad import load_vista
-try:
- import pyvista
-except ImportError:
- print("Warning: import error, module pyvista is unavailable.")
+# try:
+# import pyvista
+# except ImportError:
+# print("Warning: import error, module pyvista is unavailable.")
from quaternions import fillpositive, quat2mat, mat2quat
@@ -120,42 +121,46 @@ def loadImageFromFile(filename, pref, f_type):
print("Cannot load img/hdr pair file given!")
elif (filetype == '.v'):
- try:
- [img_data, dim, pixdim, sform_code, sform, qform_code, qform,
- voxel_res, quatern, qoffset, column_vec, row_vec, slice_vec,
- index_origin, ca, cp, extent] = pyvista.loadVImage(filename)
- hdr = {'pixdim': pixdim}
- if len(pixdim) == 1: # version 1 .v file
- pixdim.append(voxel_res[1])
- pixdim.append(voxel_res[2])
- pixdim.append(voxel_res[3])
- # Build affine from old version header info
- affine = np.eye(4)
- affine[0, 0:3] = column_vec
- affine[1, 0:3] = row_vec
- affine[2, 0:3] = slice_vec
- affine[0:3, 3] = np.transpose(index_origin)
- # if affine is still build incorrectly, use identity
- if np.linalg.det(affine[0:2, 0:2]) == 0:
- affine = np.identity(4)
- image = Nifti2Image(img_data.astype(np.float32), affine)
- else: # version 2 .v file
- # sform, or qform, and if not, then fall-back affine
- if sform_code != 0:
- affine = sform
- elif qform_code != 0:
- qform_affine = calculateQFormAffine(pixdim, qform_code, qform)
- affine = qform_affine
- else:
- affine = shape_zoom_affine(img_data.shape, pixdim[1:4], False)
- image = Nifti2Image(img_data.astype(np.float32), affine)
- if sform_code != 0:
- image.set_sform(sform)
- if qform_code != 0:
- qform_affine = calculateQFormAffine(pixdim, qform_code, qform)
- image.set_qform(qform_affine)
- except RuntimeError:
- print("Cannot load .v file given!")
+ image = load_vista(filename)
+ hdr = image.header
+
+ #old way of loading vista file!
+ # try:
+ # [img_data, dim, pixdim, sform_code, sform, qform_code, qform,
+ # voxel_res, quatern, qoffset, column_vec, row_vec, slice_vec,
+ # index_origin, ca, cp, extent] = pyvista.loadVImage(filename)
+ # hdr = {'pixdim': pixdim}
+ # if len(pixdim) == 1: # version 1 .v file
+ # pixdim.append(voxel_res[1])
+ # pixdim.append(voxel_res[2])
+ # pixdim.append(voxel_res[3])
+ # # Build affine from old version header info
+ # affine = np.eye(4)
+ # affine[0, 0:3] = column_vec
+ # affine[1, 0:3] = row_vec
+ # affine[2, 0:3] = slice_vec
+ # affine[0:3, 3] = np.transpose(index_origin)
+ # # if affine is still build incorrectly, use identity
+ # if np.linalg.det(affine[0:2, 0:2]) == 0:
+ # affine = np.identity(4)
+ # image = Nifti2Image(img_data.astype(np.float32), affine)
+ # else: # version 2 .v file
+ # # sform, or qform, and if not, then fall-back affine
+ # if sform_code != 0:
+ # affine = sform
+ # elif qform_code != 0:
+ # qform_affine = calculateQFormAffine(pixdim, qform_code, qform)
+ # affine = qform_affine
+ # else:
+ # affine = shape_zoom_affine(img_data.shape, pixdim[1:4], False)
+ # image = Nifti2Image(img_data.astype(np.float32), affine)
+ # if sform_code != 0:
+ # image.set_sform(sform)
+ # if qform_code != 0:
+ # qform_affine = calculateQFormAffine(pixdim, qform_code, qform)
+ # image.set_qform(qform_affine)
+ # except RuntimeError:
+ # print("Cannot load .v file given!")
# if data contain NaNs convert to zero
if np.isnan(image.get_data()).any():
diff --git a/vviewer/vviewer.py b/vviewer/vviewer.py
index 3a693ec1260ba9dd6a79ba77e18c852a0efb473b..bbb831a6f714dd9f1bdc95b4207c78dc3fddd746 100755
--- a/vviewer/vviewer.py
+++ b/vviewer/vviewer.py
@@ -442,27 +442,29 @@ class vviewer(QtGui.QMainWindow):
# Crosshair intensity label showing the value of the current image
# label 'cross'
- self.cross_int_label = QtGui.QLabel('Cross:')
- self.cross_int_label.setAlignment(
- QtCore.Qt.AlignVCenter | QtCore.Qt.AlignRight)
- self.l.addWidget(self.cross_int_label, 6, self.listoffset+2, 1, 2)
+ # self.cross_int_label = QtGui.QLabel('Cross:')
+ # self.cross_int_label.setAlignment(
+ # QtCore.Qt.AlignVCenter | QtCore.Qt.AlignRight)
+ # self.l.addWidget(self.cross_int_label, 6, self.listoffset+2, 1, 2)
+
+
# label for actual value
self.intensity_label = QtGui.QLabel('nan')
self.intensity_label.setAlignment(
QtCore.Qt.AlignVCenter | QtCore.Qt.AlignLeft)
# self.l.addWidget(self.intensity_label, 6, self.listoffset+4, 1, 8)
- self.l.addWidget(self.intensity_label, 6, self.listoffset+4, 1, 8)
+ self.l.addWidget(self.intensity_label, 6, self.listoffset+2, 1, 8)
# Label showing the values for mouse position
# label 'cursor'
- self.cursor_int_label = QtGui.QLabel('Cursor:')
- self.cursor_int_label.setAlignment(
- QtCore.Qt.AlignVCenter | QtCore.Qt.AlignRight)
- self.l.addWidget(self.cursor_int_label, 7, self.listoffset+2, 1, 2)
- self.intensity_lbl_cursor = QtGui.QLabel('nan')
- self.intensity_lbl_cursor.setAlignment(
- QtCore.Qt.AlignVCenter | QtCore.Qt.AlignLeft)
- self.l.addWidget(self.intensity_lbl_cursor, 7, self.listoffset+4, 1, 10)
+ # self.cursor_int_label = QtGui.QLabel('Cursor:')
+ # self.cursor_int_label.setAlignment(
+ # QtCore.Qt.AlignVCenter | QtCore.Qt.AlignRight)
+ # self.l.addWidget(self.cursor_int_label, 7, self.listoffset+2, 1, 2)
+ # self.intensity_lbl_cursor = QtGui.QLabel('nan')
+ # self.intensity_lbl_cursor.setAlignment(
+ # QtCore.Qt.AlignVCenter | QtCore.Qt.AlignLeft)
+ # self.l.addWidget(self.intensity_lbl_cursor, 7, self.listoffset+4, 1, 10)
## For playing time series movies ##
# fast backward button
@@ -1542,7 +1544,7 @@ class vviewer(QtGui.QMainWindow):
intensity = ("%.6f" %self.images[index].getIntensity())
else:
intensity = "nan"
- self.intensity_lbl_cursor.setText(intensity)
+ # self.intensity_lbl_cursor.setText(intensity)
if self.value_window.isVisible():
if coords_valid:
names = "<b>filename</b><br>"
@@ -1594,7 +1596,18 @@ class vviewer(QtGui.QMainWindow):
#intensity = (
# str(np.round(self.images[index].getIntensity(),3)))
# intensity = (str(self.images[index].getIntensity()))
- intensity = ("%.6f" %self.images[index].getIntensity())
+
+
+ # intensity = ("%.6f" %self.images[index].getIntensity())
+ #jjj change
+ nmb_images = len(self.images)
+ str_intens = ""
+ for i in range(nmb_images):
+ if i > 0:
+ str_intens += "\n"
+ str_intens += "%.3f" %self.images[i].getIntensity()
+ str_intens += " (%s)" %str(self.imagelist.item(i).text()).split(".")[0]
+ intensity = (str_intens)
self.intensity_label.setText(intensity)
# update ValueWindow