def __init__(self):
QtGui.QMainWindow.__init__(self)
self.ui = Ui_MainWindow()
self.ui.setupUi(self)
self.setupSignals()
def __init__(self):
QtGui.QMainWindow.__init__(self)
self.ui = Ui_MainWindow()
self.ui.setupUi(self)
self.setupSignals()
class Main(QtGui.QMainWindow):
def __init__(self):
QtGui.QMainWindow.__init__(self)
self.ui = Ui_MainWindow()
self.ui.setupUi(self)
self.setupSignals()
def setupSignals(self):
self.ui.btnReset.clicked.connect(self.reset)
self.ui.btnApply.clicked.connect(self.setWorkingImg)
self.ui.btnPicamSnap.clicked.connect(self.picamSnap)
# radio buttons under orig and active images
self.ui.rbOrigWorking.toggled.connect(self.showOrigImage)
self.ui.rbOrigOrig.toggled.connect(self.showOrigImage)
self.ui.rbOrigTemplate.toggled.connect(self.showOrigImage)
self.ui.rbActiveActive.toggled.connect(self.showActiveImage)
self.ui.rbActiveOriginal.toggled.connect(self.showActiveImage)
# functions under menu bar
self.ui.actionOpen.triggered.connect(self.fileOpen)
self.ui.actionLoad_template.triggered.connect(self.templateOpen)
self.ui.actionSave_Image.triggered.connect(self.fileSave)
# functions under basic tab
self.ui.btnRotate90.clicked.connect(lambda: self.rotateImg(90))
self.ui.btnRotate180.clicked.connect(lambda: self.rotateImg(180))
self.ui.btnRotate270.clicked.connect(lambda: self.rotateImg(270))
self.ui.sbRotateX.valueChanged.connect(
lambda: self.rotateImg(self.ui.sbRotateX.value()))
# functions under Region of Interest (ROI)
self.ui.sliderROIBottom.valueChanged.connect(self.roiSelect)
self.ui.sliderROITop.valueChanged.connect(self.roiSelect)
self.ui.sliderROILeft.valueChanged.connect(self.roiSelect)
self.ui.sliderROIRight.valueChanged.connect(self.roiSelect)
# color operations
self.ui.listGreyscale.currentItemChanged.connect(self.toGreyscale)
self.ui.sbLH.valueChanged.connect(self.hsvMask)
self.ui.sbLS.valueChanged.connect(self.hsvMask)
self.ui.sbLV.valueChanged.connect(self.hsvMask)
self.ui.sbUH.valueChanged.connect(self.hsvMask)
self.ui.sbUS.valueChanged.connect(self.hsvMask)
self.ui.sbUV.valueChanged.connect(self.hsvMask)
# greyscale operations
self.ui.btnCannyEdges.clicked.connect(self.auto_canny)
self.ui.btnEqualizeHist.clicked.connect(self.equalizeHist)
self.ui.btnCLAHE.clicked.connect(self.clahe)
self.ui.btnBlurAveraging.clicked.connect(self.blurAveraging)
self.ui.btnBlurGaussian.clicked.connect(self.blurGaussian)
self.ui.btnBlurMedian.clicked.connect(self.blurMedian)
self.ui.btnFilterBilateral.clicked.connect(self.filterBilateral)
self.ui.btnThreshGlobal.clicked.connect(self.threshGlobal)
self.ui.btnThreshGaussian.clicked.connect(self.threshGaussian)
self.ui.btnThreshMean.clicked.connect(self.threshMean)
# binary operations
self.ui.btnErode.clicked.connect(self.erosion)
self.ui.btnDilate.clicked.connect(self.dilation)
self.ui.btnOpen.clicked.connect(self.opening)
self.ui.btnClose.clicked.connect(self.closing)
# find contours operations
self.ui.btnFindRectangle.clicked.connect(self.findRectangle)
self.ui.btnFindMinRectangle.clicked.connect(self.findMinAreaRectangle)
self.ui.btnFindMinCircle.clicked.connect(self.findMinCircle)
self.ui.btnFitEllipse.clicked.connect(self.fitEllipse)
# feature detection operations
self.ui.btnTemplateFind.clicked.connect(self.matchTemplate)
self.ui.btnFeatureDetect.clicked.connect(self.templateSURF)
def mat2Pixmap(self, cv2Image, lbl):
# convert image from cv mat type to qt pixmap
qtImage = self.toQImage(cv2Image)
# convert qt image to pixmap
qtPixmap = QtGui.QPixmap.fromImage(qtImage)
# scale image to fit inside window
qtPixmap = qtPixmap.scaled(lbl.size(), QtCore.Qt.KeepAspectRatio)
return qtPixmap
def toQImage(self, im, copy=False):
# converts the cv mat type image to a Qimage type
gray_color_table = [QtGui.qRgb(i, i, i) for i in range(256)]
if im is None:
return QtGui.QImage()
if im.dtype == np.uint8:
if len(im.shape) == 2:
qim = QtGui.QImage(im.data, im.shape[1], im.shape[0],
im.strides[0], QtGui.QImage.Format_Indexed8)
qim.setColorTable(gray_color_table)
return qim.copy() if copy else qim
elif len(im.shape) == 3:
if im.shape[2] == 3:
qim = QtGui.QImage(im.data, im.shape[1], im.shape[0], im.strides[0], QtGui.QImage.Format_RGB888)\
.rgbSwapped()
return qim.copy() if copy else qim
elif im.shape[2] == 4:
qim = QtGui.QImage(im.data, im.shape[1], im.shape[0], im.strides[0], QtGui.QImage.Format_ARGB32)\
.rgbSwapped()
return qim.copy() if copy else qim
def setWorkingImg(self):
global workingImage
print("updating working image")
workingImage = activeImage
self.updateWorkingImg()
def updateOrigImg(self):
origImagePx = self.mat2Pixmap(origImage, self.ui.lblOrigOrig)
self.ui.lblOrigOrig.setPixmap(origImagePx)
self.ui.lblActiveOrig.setPixmap(origImagePx)
def updateActiveImg(self):
# create QT pixmap datatype for QT to display
activeImagePx = self.mat2Pixmap(activeImage, self.ui.lblActiveActive)
# Putting pixmap image into QT label container
self.ui.lblActiveActive.setPixmap(activeImagePx)
def updateWorkingImg(self):
workingImagePx = self.mat2Pixmap(workingImage, self.ui.lblOrigWorking)
self.ui.lblOrigWorking.setPixmap(workingImagePx)
def updateTemplateImg(self):
templateImagePx = self.mat2Pixmap(templateImage,
self.ui.lblActiveActive)
self.ui.lblOrigTemplate.setPixmap(templateImagePx)
def fileOpen(self):
# load image file, set image to origImage, activeImage, and workingImage
file_name = QtGui.QFileDialog.getOpenFileName(
self, 'Open Photo', '/home/pi/Desktop/',
'images(*.png *.jpg *.jpeg)')
global origImage
origImage = cv2.imread(str(file_name))
global workingImage
workingImage = origImage
global activeImage
activeImage = origImage
self.updateActiveImg()
self.updateWorkingImg()
self.updateOrigImg()
def templateOpen(self):
# load image file, set image to origImage, activeImage, and workingImage
file_name = QtGui.QFileDialog.getOpenFileName(
self, 'Open Photo', '/home/pi/Desktop/',
'images(*.png *.jpg *.jpeg)')
global templateImage
templateImage = cv2.imread(str(file_name), 0)
self.updateTemplateImg()
def fileSave(self):
# save the image currently in the workingImage variable
file_name = QtGui.QFileDialog.getSaveFileName(self, 'Save File', '',
'images(*.png')
file_name = str(file_name) + ".png"
cv2.imwrite(file_name, workingImage)
def rotateImg(self, angle):
(h, w) = workingImage.shape[:2]
center = (w / 2, h / 2)
M = cv2.getRotationMatrix2D(center, angle, 1.0)
global activeImage
activeImage = cv2.warpAffine(workingImage, M, (w, h))
self.updateActiveImg()
def roiSelect(self):
global activeImage
if self.ui.rbROIEnable.isChecked():
(h, w) = workingImage.shape[:2]
x1 = int(self.ui.sliderROILeft.value() / 100.0 * w)
x2 = int((1 - self.ui.sliderROIRight.value() / 100.0) * w)
y1 = int(self.ui.sliderROITop.value() / 100.0 * h)
y2 = int((1 - self.ui.sliderROIBottom.value() / 100.0) * h)
print(x1, x2, y1, y2)
if self.ui.rbROICrop.isChecked():
activeImage = workingImage[y1:y2, x1:x2].copy()
self.updateActiveImg()
else:
pass
def reset(self):
global activeImage
activeImage = origImage
global workingImage
workingImage = origImage
self.updateActiveImg()
self.updateWorkingImg()
def showOrigImage(self):
if self.ui.rbOrigWorking.isChecked():
self.ui.stackOrig.setCurrentIndex(0)
elif self.ui.rbOrigOrig.isChecked():
self.ui.stackOrig.setCurrentIndex(1)
elif self.ui.rbOrigTemplate.isChecked():
self.ui.stackOrig.setCurrentIndex(2)
def showActiveImage(self):
if self.ui.rbActiveActive.isChecked():
self.ui.stackActive.setCurrentIndex(0)
elif self.ui.rbActiveOriginal.isChecked():
self.ui.stackActive.setCurrentIndex(1)
def toGreyscale(self, item):
itemtext = str(item.text())
global activeImage
# convert image to greyscale through simple methods
if itemtext == "Simple Greyscale":
activeImage = cv2.cvtColor(workingImage, cv2.COLOR_BGR2GRAY)
# extract red plane and convert to greyscale
if itemtext in ("Blue", "Green", "Red"):
# activeImage = workingImage[:, :, 0]
b, g, r = cv2.split(workingImage)
if itemtext == "Red":
activeImage = r
if itemtext == "Green":
activeImage = g
if itemtext == "Blue":
activeImage = b
# extract Hue, Saturation, Value
if itemtext in ("Hue", "Saturation", "Value"):
hsv = cv2.cvtColor(workingImage, cv2.COLOR_BGR2HSV)
h, s, v = cv2.split(hsv)
if itemtext == "Hue":
activeImage = h
if itemtext == "Saturation":
activeImage = s
if itemtext == "Value":
activeImage = v
# extract Lightness plane from HSL
if itemtext == "Lightness":
hsl = cv2.cvtColor(workingImage, cv2.COLOR_BGR2HSL)
h, s, l = cv2.split(hsl)
activeImage = l
# exract CIELAB L* a* b*
if itemtext in ("L*", "a*", "b*"):
Lab = cv2.cvtColor(workingImage, cv2.COLOR_BGR2HSV)
L, a, b = cv2.split(Lab)
if itemtext == "L*":
activeImage = L
if itemtext == "a*":
activeImage = a
if itemtext == "b*":
activeImage = b
# update the active image
self.updateActiveImg()
def erosion(self):
# Erode the image using kernel specified
kernel = np.ones(
(self.ui.sbKernel1.value(), self.ui.sbKernel2.value()), np.uint8)
global activeImage
activeImage = cv2.erode(workingImage, kernel, iterations=1)
self.updateActiveImg()
def dilation(self):
kernel = np.ones(
(self.ui.sbKernel1.value(), self.ui.sbKernel2.value()), np.uint8)
global activeImage
activeImage = cv2.dilate(workingImage, kernel, iterations=1)
self.updateActiveImg()
def opening(self):
kernel = np.ones(
(self.ui.sbKernel1.value(), self.ui.sbKernel2.value()), np.uint8)
global activeImage
activeImage = cv2.morphologyEx(workingImage, cv2.MORPH_OPEN, kernel)
self.updateActiveImg()
def closing(self):
kernel = np.ones(
(self.ui.sbKernel1.value(), self.ui.sbKernel2.value()), np.uint8)
global activeImage
activeImage = cv2.morphologyEx(workingImage, cv2.MORPH_CLOSE, kernel)
self.updateActiveImg()
# Histogram equalization
def equalizeHist(self):
global activeImage
activeImage = cv2.equalizeHist(workingImage)
self.updateActiveImg()
# Contrast Limited Adaptive Histogram Equalization
def clahe(self):
claheobj = cv2.createCLAHE(clipLimit=2.0, tileGridSize=(8, 8))
global activeImage
activeImage = claheobj.apply(workingImage)
self.updateActiveImg()
def blurAveraging(self):
kernel = (self.ui.sbKernel1.value(), self.ui.sbKernel2.value())
global activeImage
activeImage = cv2.blur(workingImage, kernel)
self.updateActiveImg()
def blurGaussian(self):
kernel = (self.ui.sbKernel1.value(), self.ui.sbKernel2.value())
global activeImage
activeImage = cv2.GaussianBlur(workingImage, kernel, 0)
self.updateActiveImg()
def blurMedian(self):
kernel = self.ui.sbKernel1.value()
global activeImage
activeImage = cv2.medianBlur(workingImage, kernel)
self.updateActiveImg()
def filterBilateral(self):
global activeImage
activeImage = cv2.bilateralFilter(workingImage, 9, 75, 75)
self.updateActiveImg()
def getThreshType(self, index):
threshType = cv2.THRESH_BINARY
if index == "BINARY":
threshType = cv2.THRESH_BINARY
if index == "BINARY_INV":
threshType = cv2.THRESH_BINARY_INV
if index == "TOZERO":
threshType = cv2.THRESH_TOZERO
if index == "TRUNC":
threshType = cv2.THRESH_TRUNC
if index == "TOZERO_INV":
threshType = cv2.THRESH_TOZERO_INV
return (threshType)
def threshGlobal(self):
threshType = self.getThreshType(self.ui.cbThreshType.currentText())
global activeImage
ret, activeImage = cv2.threshold(workingImage,
self.ui.sbThreshThresh.value(),
self.ui.sbThreshMaxval.value(),
threshType)
self.updateActiveImg()
def threshMean(self):
threshType = self.getThreshType(self.ui.cbThreshType.currentText())
global activeImage
activeImage = cv2.adaptiveThreshold(workingImage,
self.ui.sbThreshMaxval.value(),
cv2.ADAPTIVE_THRESH_MEAN_C,
threshType, 11, 2)
self.updateActiveImg()
def threshGaussian(self):
threshType = self.getThreshType(self.ui.cbThreshType.currentText())
global activeImage
activeImage = cv2.adaptiveThreshold(workingImage,
self.ui.sbThreshMaxval.value(),
cv2.ADAPTIVE_THRESH_GAUSSIAN_C,
threshType, 11, 2)
self.updateActiveImg()
def auto_canny(self, sigma=0.33):
# compute the median of the single channel pixel intensities
v = np.median(workingImage)
# apply automatic Canny edge detection using the computed median
lower = int(max(0, (1.0 - sigma) * v))
upper = int(min(255, (1.0 + sigma) * v))
global activeImage
activeImage = cv2.Canny(activeImage, lower, upper)
self.updateActiveImg()
def hsvMask(self):
# Convert BGR to HSV
hsv = cv2.cvtColor(workingImage, cv2.COLOR_BGR2HSV)
# define range of blue color in HSV
lower_blue = np.array(
[self.ui.sbLH.value(),
self.ui.sbLS.value(),
self.ui.sbLV.value()])
upper_blue = np.array(
[self.ui.sbUH.value(),
self.ui.sbUS.value(),
self.ui.sbUV.value()])
# Threshold the HSV image to get only blue colors
mask = cv2.inRange(hsv, lower_blue, upper_blue)
# Bitwise-AND mask and original image
global activeImage
# activeImage = cv2.bitwise_and(workingImage,workingImage, mask= mask) #display color image
# Uncomment above and comment the below statement to see color instead of binary as active image
activeImage = mask # display binarized image
# Show the masked image
self.updateActiveImg()
def findContours(self):
# Find the contours
imgContours, contours, hierarchy = cv2.findContours(
workingImage.copy(), cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
contours = sorted(contours, key=cv2.contourArea, reverse=True)[:10]
return contours
def findRectangle(self):
contours = self.findContours()
x, y, w, h = cv2.boundingRect(contours[0])
global activeImage
activeImage = cv2.rectangle(origImage.copy(), (x, y), (x + w, y + h),
(0, 0, 255), 4)
self.updateActiveImg()
def findMinAreaRectangle(self):
contours = self.findContours()
rect = cv2.minAreaRect(contours[0])
box = cv2.boxPoints(rect)
box = np.int0(box)
global activeImage
activeImage = cv2.drawContours(origImage.copy(), [box], 0, (0, 0, 255),
4)
self.updateActiveImg()
def findMinCircle(self):
contours = self.findContours()
(x, y), radius = cv2.minEnclosingCircle(contours[0])
center = (int(x), int(y))
radius = int(radius)
global activeImage
activeImage = cv2.circle(origImage.copy(), center, radius, (0, 255, 0),
4)
self.updateActiveImg()
def fitEllipse(self):
contours = self.findContours()
ellipse = cv2.fitEllipse(contours[0])
global activeImage
activeImage = cv2.ellipse(origImage.copy(), ellipse, (0, 255, 0), 4)
self.updateActiveImg()
def matchTemplate(self):
# sets method equal to what is in the Template Method combo box
method = 0
if self.ui.cbTemplateMethod.currentText() == "SQDIFF":
method = cv2.TM_SQDIFF
elif self.ui.cbTemplateMethod.currentText() == "SQDIFF_NORMED":
method = cv2.TM_SQDIFF_NORMED
elif self.ui.cbTemplateMethod.currentText() == "CCORR":
method = cv2.TM_CCORR
elif self.ui.cbTemplateMethod.currentText() == "CCORR_NORMED":
method = cv2.TM_CCORR_NORMED
elif self.ui.cbTemplateMethod.currentText() == "CCOEFF":
method = cv2.TM_CCOEFF
elif self.ui.cbTemplateMethod.currentText() == "CCOEFF_NORMED":
method = cv2.TM_CCOEFF_NORMED
# perform the cv2.matchTemplate function and store output to 'result'
result = cv2.matchTemplate(workingImage, templateImage, method)
# find the minimum and maximum match values as well as their locations on the photo
min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(result)
# If the method is TM_SQDIFF or TM_SQDIFF_NORMED, take minimum
if method in [cv2.TM_SQDIFF, cv2.TM_SQDIFF_NORMED]:
top_left = min_loc
else: # all other methods
top_left = max_loc
# height and width of template photo
h, w = templateImage.shape[:2]
# draw a rectangle on the active image
bottom_right = (top_left[0] + w, top_left[1] + h)
global activeImage
# create a copy of the workingImage object since cv2.rectangle modifies the object directly
# omitting the following line causes cv2.rectangle draw a rectangle on all references to the image object
activeImage = workingImage.copy()
cv2.rectangle(activeImage, top_left, bottom_right, 255, 4)
self.updateActiveImg()
def templateSURF(self):
if 'surf' not in globals():
global surf
print("creating surf component")
surf = cv2.xfeatures2d.SURF_create(400)
# find the keypoints and descriptors with SURF
kpT, desT = surf.detectAndCompute(templateImage, None)
kpA, desA = surf.detectAndCompute(activeImage, None)
FLANN_INDEX_KDTREE = 0
index_params = dict(algorithm=FLANN_INDEX_KDTREE, trees=5)
search_params = dict(checks=50)
flann = cv2.FlannBasedMatcher(index_params, search_params)
matches = flann.knnMatch(desT, desA, k=2)
# store all the good matches as per Lowe's ratio test
good = []
for m, n in matches:
if m.distance < 0.7 * n.distance:
good.append(m)
MIN_MATCH_COUNT = self.ui.sbFeatureMinCnt.value()
print(len(good))
if len(good) > MIN_MATCH_COUNT:
src_pts = np.float32([kpT[m.queryIdx].pt
for m in good]).reshape(-1, 1, 2)
dst_pts = np.float32([kpA[m.trainIdx].pt
for m in good]).reshape(-1, 1, 2)
M, mask = cv2.findHomography(src_pts, dst_pts, cv2.RANSAC, 5.0)
matchesMask = mask.ravel().tolist()
h, w = activeImage.shape
pts = np.float32([[0, 0], [0, h - 1], [w - 1, h - 1],
[w - 1, 0]]).reshape(-1, 1, 2)
dst = cv2.perspectiveTransform(pts, M)
global activeImage
activeImage = activeImage.copy()
activeImage = cv2.polylines(activeImage, [np.int32(dst)], True,
255, 3, cv2.LINE_AA)
else:
print "Not enough matches are found -%d/%d" % (len(good),
MIN_MATCH_COUNT)
matchesMask = None
draw_params = dict(
matchColor=(0, 255, 0), # draw matches in green color
singlePointColor=None,
matchesMask=matchesMask, # draw only inliers
flags=2)
matchedImage = cv2.drawMatches(templateImage, kpT, activeImage, kpA,
good, None, **draw_params)
cv2.imshow("matched image", matchedImage)
def picamSnap(self):
# initialize the camera and grab a reference to the raw camera capture
# set camera resolution to what is selected in the combo box
camera.resolution = map(
int,
str(self.ui.cbPicamResolution.currentText()).split("x"))
rawCapture = PiRGBArray(camera)
# allow the camera to warmup
time.sleep(0.1)
# grab an image from the camera
camera.capture(rawCapture, format="bgr")
image = rawCapture.array
global activeImage
activeImage = image
global workingImage
workingImage = image
global origImage
origImage = image
self.updateActiveImg()
self.updateWorkingImg()
self.updateOrigImg()
class Main(QtGui.QMainWindow):
def __init__(self):
QtGui.QMainWindow.__init__(self)
self.ui = Ui_MainWindow()
self.ui.setupUi(self)
self.setupSignals()
def setupSignals(self):
self.ui.btnReset.clicked.connect(self.reset)
self.ui.btnApply.clicked.connect(self.setWorkingImg)
self.ui.btnPicamSnap.clicked.connect(self.picamSnap)
# radio buttons under orig and active images
self.ui.rbOrigWorking.toggled.connect(self.showOrigImage)
self.ui.rbOrigOrig.toggled.connect(self.showOrigImage)
self.ui.rbOrigTemplate.toggled.connect(self.showOrigImage)
self.ui.rbActiveActive.toggled.connect(self.showActiveImage)
self.ui.rbActiveOriginal.toggled.connect(self.showActiveImage)
# functions under menu bar
self.ui.actionOpen.triggered.connect(self.fileOpen)
self.ui.actionLoad_template.triggered.connect(self.templateOpen)
self.ui.actionSave_Image.triggered.connect(self.fileSave)
# functions under basic tab
self.ui.btnRotate90.clicked.connect(lambda: self.rotateImg(90))
self.ui.btnRotate180.clicked.connect(lambda: self.rotateImg(180))
self.ui.btnRotate270.clicked.connect(lambda: self.rotateImg(270))
self.ui.sbRotateX.valueChanged.connect(lambda: self.rotateImg(self.ui.sbRotateX.value()))
# functions under Region of Interest (ROI)
self.ui.sliderROIBottom.valueChanged.connect(self.roiSelect)
self.ui.sliderROITop.valueChanged.connect(self.roiSelect)
self.ui.sliderROILeft.valueChanged.connect(self.roiSelect)
self.ui.sliderROIRight.valueChanged.connect(self.roiSelect)
# color operations
self.ui.listGreyscale.currentItemChanged.connect(self.toGreyscale)
self.ui.sbLH.valueChanged.connect(self.hsvMask)
self.ui.sbLS.valueChanged.connect(self.hsvMask)
self.ui.sbLV.valueChanged.connect(self.hsvMask)
self.ui.sbUH.valueChanged.connect(self.hsvMask)
self.ui.sbUS.valueChanged.connect(self.hsvMask)
self.ui.sbUV.valueChanged.connect(self.hsvMask)
# greyscale operations
self.ui.btnCannyEdges.clicked.connect(self.auto_canny)
self.ui.btnEqualizeHist.clicked.connect(self.equalizeHist)
self.ui.btnCLAHE.clicked.connect(self.clahe)
self.ui.btnBlurAveraging.clicked.connect(self.blurAveraging)
self.ui.btnBlurGaussian.clicked.connect(self.blurGaussian)
self.ui.btnBlurMedian.clicked.connect(self.blurMedian)
self.ui.btnFilterBilateral.clicked.connect(self.filterBilateral)
self.ui.btnThreshGlobal.clicked.connect(self.threshGlobal)
self.ui.btnThreshGaussian.clicked.connect(self.threshGaussian)
self.ui.btnThreshMean.clicked.connect(self.threshMean)
# binary operations
self.ui.btnErode.clicked.connect(self.erosion)
self.ui.btnDilate.clicked.connect(self.dilation)
self.ui.btnOpen.clicked.connect(self.opening)
self.ui.btnClose.clicked.connect(self.closing)
# find contours operations
self.ui.btnFindRectangle.clicked.connect(self.findRectangle)
self.ui.btnFindMinRectangle.clicked.connect(self.findMinAreaRectangle)
self.ui.btnFindMinCircle.clicked.connect(self.findMinCircle)
self.ui.btnFitEllipse.clicked.connect(self.fitEllipse)
# feature detection operations
self.ui.btnTemplateFind.clicked.connect(self.matchTemplate)
self.ui.btnFeatureDetect.clicked.connect(self.templateSURF)
def mat2Pixmap(self, cv2Image, lbl):
# convert image from cv mat type to qt pixmap
qtImage = self.toQImage(cv2Image)
# convert qt image to pixmap
qtPixmap = QtGui.QPixmap.fromImage(qtImage)
# scale image to fit inside window
qtPixmap = qtPixmap.scaled(lbl.size(), QtCore.Qt.KeepAspectRatio)
return qtPixmap
def toQImage(self, im, copy=False):
# converts the cv mat type image to a Qimage type
gray_color_table = [QtGui.qRgb(i, i, i) for i in range(256)]
if im is None:
return QtGui.QImage()
if im.dtype == np.uint8:
if len(im.shape) == 2:
qim = QtGui.QImage(im.data, im.shape[1], im.shape[0], im.strides[0], QtGui.QImage.Format_Indexed8)
qim.setColorTable(gray_color_table)
return qim.copy() if copy else qim
elif len(im.shape) == 3:
if im.shape[2] == 3:
qim = QtGui.QImage(im.data, im.shape[1], im.shape[0], im.strides[0], QtGui.QImage.Format_RGB888)\
.rgbSwapped();
return qim.copy() if copy else qim
elif im.shape[2] == 4:
qim = QtGui.QImage(im.data, im.shape[1], im.shape[0], im.strides[0], QtGui.QImage.Format_ARGB32)\
.rgbSwapped();
return qim.copy() if copy else qim
def setWorkingImg(self):
global workingImage
print ("updating working image")
workingImage = activeImage
self.updateWorkingImg()
def updateOrigImg(self):
origImagePx = self.mat2Pixmap(origImage, self.ui.lblOrigOrig)
self.ui.lblOrigOrig.setPixmap(origImagePx)
self.ui.lblActiveOrig.setPixmap(origImagePx)
def updateActiveImg(self):
# create QT pixmap datatype for QT to display
activeImagePx = self.mat2Pixmap(activeImage, self.ui.lblActiveActive)
# Putting pixmap image into QT label container
self.ui.lblActiveActive.setPixmap(activeImagePx)
def updateWorkingImg(self):
workingImagePx = self.mat2Pixmap(workingImage, self.ui.lblOrigWorking)
self.ui.lblOrigWorking.setPixmap(workingImagePx)
def updateTemplateImg(self):
templateImagePx = self.mat2Pixmap(templateImage, self.ui.lblActiveActive)
self.ui.lblOrigTemplate.setPixmap(templateImagePx)
def fileOpen(self):
# load image file, set image to origImage, activeImage, and workingImage
file_name = QtGui.QFileDialog.getOpenFileName(self, 'Open Photo', '/home/pi/Desktop/', 'images(*.png *.jpg *.jpeg)')
global origImage
origImage = cv2.imread(str(file_name))
global workingImage
workingImage = origImage
global activeImage
activeImage = origImage
self.updateActiveImg()
self.updateWorkingImg()
self.updateOrigImg()
def templateOpen(self):
# load image file, set image to origImage, activeImage, and workingImage
file_name = QtGui.QFileDialog.getOpenFileName(self, 'Open Photo', '/home/pi/Desktop/', 'images(*.png *.jpg *.jpeg)')
global templateImage
templateImage = cv2.imread(str(file_name), 0)
self.updateTemplateImg()
def fileSave(self):
# save the image currently in the workingImage variable
file_name = QtGui.QFileDialog.getSaveFileName(self, 'Save File', '', 'images(*.png')
file_name = str(file_name) + ".png"
cv2.imwrite(file_name, workingImage)
def rotateImg(self, angle):
(h, w) = workingImage.shape[:2]
center = (w/2, h/2)
M = cv2.getRotationMatrix2D(center, angle, 1.0)
global activeImage
activeImage = cv2.warpAffine(workingImage, M, (w, h))
self.updateActiveImg()
def roiSelect(self):
global activeImage
if self.ui.rbROIEnable.isChecked():
(h, w) = workingImage.shape[:2]
x1 = int(self.ui.sliderROILeft.value()/100.0 * w)
x2 = int((1 - self.ui.sliderROIRight.value()/100.0) * w)
y1 = int(self.ui.sliderROITop.value()/100.0 * h)
y2 = int((1 - self.ui.sliderROIBottom.value()/100.0) * h)
print(x1, x2, y1, y2)
if self.ui.rbROICrop.isChecked():
activeImage = workingImage[y1:y2, x1:x2].copy()
self.updateActiveImg()
else:
pass
def reset(self):
global activeImage
activeImage = origImage
global workingImage
workingImage = origImage
self.updateActiveImg()
self.updateWorkingImg()
def showOrigImage(self):
if self.ui.rbOrigWorking.isChecked():
self.ui.stackOrig.setCurrentIndex(0)
elif self.ui.rbOrigOrig.isChecked():
self.ui.stackOrig.setCurrentIndex(1)
elif self.ui.rbOrigTemplate.isChecked():
self.ui.stackOrig.setCurrentIndex(2)
def showActiveImage(self):
if self.ui.rbActiveActive.isChecked():
self.ui.stackActive.setCurrentIndex(0)
elif self.ui.rbActiveOriginal.isChecked():
self.ui.stackActive.setCurrentIndex(1)
def toGreyscale(self, item):
itemtext = str(item.text())
global activeImage
# convert image to greyscale through simple methods
if itemtext == "Simple Greyscale":
activeImage = cv2.cvtColor(workingImage,cv2.COLOR_BGR2GRAY)
# extract red plane and convert to greyscale
if itemtext in ("Blue", "Green", "Red"):
# activeImage = workingImage[:, :, 0]
b, g, r = cv2.split(workingImage)
if itemtext == "Red":
activeImage = r
if itemtext == "Green":
activeImage = g
if itemtext == "Blue":
activeImage = b
# extract Hue, Saturation, Value
if itemtext in ("Hue", "Saturation", "Value"):
hsv = cv2.cvtColor(workingImage, cv2.COLOR_BGR2HSV)
h, s, v = cv2.split(hsv)
if itemtext == "Hue":
activeImage = h
if itemtext == "Saturation":
activeImage = s
if itemtext == "Value":
activeImage = v
# extract Lightness plane from HSL
if itemtext == "Lightness":
hsl = cv2.cvtColor(workingImage, cv2.COLOR_BGR2HSL)
h, s, l = cv2.split(hsl)
activeImage = l
# exract CIELAB L* a* b*
if itemtext in ("L*", "a*", "b*"):
Lab = cv2.cvtColor(workingImage, cv2.COLOR_BGR2HSV)
L, a, b = cv2.split(Lab)
if itemtext == "L*":
activeImage = L
if itemtext == "a*":
activeImage = a
if itemtext == "b*":
activeImage = b
# update the active image
self.updateActiveImg()
def erosion(self):
# Erode the image using kernel specified
kernel = np.ones((self.ui.sbKernel1.value(), self.ui.sbKernel2.value()), np.uint8)
global activeImage
activeImage = cv2.erode(workingImage, kernel, iterations=1)
self.updateActiveImg()
def dilation(self):
kernel = np.ones((self.ui.sbKernel1.value(), self.ui.sbKernel2.value()), np.uint8)
global activeImage
activeImage = cv2.dilate(workingImage, kernel, iterations=1)
self.updateActiveImg()
def opening(self):
kernel = np.ones((self.ui.sbKernel1.value(), self.ui.sbKernel2.value()), np.uint8)
global activeImage
activeImage = cv2.morphologyEx(workingImage, cv2.MORPH_OPEN, kernel)
self.updateActiveImg()
def closing(self):
kernel = np.ones((self.ui.sbKernel1.value(), self.ui.sbKernel2.value()), np.uint8)
global activeImage
activeImage = cv2.morphologyEx(workingImage, cv2.MORPH_CLOSE, kernel)
self.updateActiveImg()
# Histogram equalization
def equalizeHist(self):
global activeImage
activeImage = cv2.equalizeHist(workingImage)
self.updateActiveImg()
# Contrast Limited Adaptive Histogram Equalization
def clahe(self):
claheobj = cv2.createCLAHE(clipLimit=2.0,tileGridSize=(8, 8))
global activeImage
activeImage = claheobj.apply(workingImage)
self.updateActiveImg()
def blurAveraging(self):
kernel = (self.ui.sbKernel1.value(), self.ui.sbKernel2.value())
global activeImage
activeImage = cv2.blur(workingImage, kernel)
self.updateActiveImg()
def blurGaussian(self):
kernel = (self.ui.sbKernel1.value(), self.ui.sbKernel2.value())
global activeImage
activeImage = cv2.GaussianBlur(workingImage, kernel, 0)
self.updateActiveImg()
def blurMedian(self):
kernel = self.ui.sbKernel1.value()
global activeImage
activeImage = cv2.medianBlur(workingImage, kernel)
self.updateActiveImg()
def filterBilateral(self):
global activeImage
activeImage = cv2.bilateralFilter(workingImage, 9, 75, 75)
self.updateActiveImg()
def getThreshType(self, index):
threshType = cv2.THRESH_BINARY
if index == "BINARY":
threshType = cv2.THRESH_BINARY
if index == "BINARY_INV":
threshType = cv2.THRESH_BINARY_INV
if index == "TOZERO":
threshType = cv2.THRESH_TOZERO
if index == "TRUNC":
threshType = cv2.THRESH_TRUNC
if index == "TOZERO_INV":
threshType = cv2.THRESH_TOZERO_INV
return (threshType)
def threshGlobal(self):
threshType = self.getThreshType(self.ui.cbThreshType.currentText())
global activeImage
ret, activeImage = cv2.threshold(workingImage, self.ui.sbThreshThresh.value(),
self.ui.sbThreshMaxval.value(), threshType)
self.updateActiveImg()
def threshMean(self):
threshType = self.getThreshType(self.ui.cbThreshType.currentText())
global activeImage
activeImage = cv2.adaptiveThreshold(workingImage, self.ui.sbThreshMaxval.value(),
cv2.ADAPTIVE_THRESH_MEAN_C, threshType, 11, 2)
self.updateActiveImg()
def threshGaussian(self):
threshType = self.getThreshType(self.ui.cbThreshType.currentText())
global activeImage
activeImage = cv2.adaptiveThreshold(workingImage, self.ui.sbThreshMaxval.value(),
cv2.ADAPTIVE_THRESH_GAUSSIAN_C, threshType, 11, 2)
self.updateActiveImg()
def auto_canny(self, sigma=0.33):
# compute the median of the single channel pixel intensities
v = np.median(workingImage)
# apply automatic Canny edge detection using the computed median
lower = int(max(0, (1.0 - sigma) * v))
upper = int(min(255, (1.0 + sigma) * v))
global activeImage
activeImage = cv2.Canny(activeImage, lower, upper)
self.updateActiveImg()
def hsvMask(self):
# Convert BGR to HSV
hsv = cv2.cvtColor(workingImage, cv2.COLOR_BGR2HSV)
# define range of blue color in HSV
lower_blue = np.array([self.ui.sbLH.value(), self.ui.sbLS.value(), self.ui.sbLV.value()])
upper_blue = np.array([self.ui.sbUH.value(), self.ui.sbUS.value(), self.ui.sbUV.value()])
# Threshold the HSV image to get only blue colors
mask = cv2.inRange(hsv, lower_blue, upper_blue)
# Bitwise-AND mask and original image
global activeImage
# activeImage = cv2.bitwise_and(workingImage,workingImage, mask= mask) #display color image
# Uncomment above and comment the below statement to see color instead of binary as active image
activeImage = mask # display binarized image
# Show the masked image
self.updateActiveImg()
def findContours(self):
# Find the contours
imgContours, contours, hierarchy = cv2.findContours(workingImage.copy(), cv2.RETR_TREE, cv2.CHAIN_APPROX_SIMPLE)
contours = sorted(contours, key=cv2.contourArea, reverse=True)[:10]
return contours
def findRectangle(self):
contours = self.findContours()
x, y, w, h = cv2.boundingRect(contours[0])
global activeImage
activeImage = cv2.rectangle(origImage.copy(), (x, y), (x+w, y+h), (0, 0, 255), 4)
self.updateActiveImg()
def findMinAreaRectangle(self):
contours = self.findContours()
rect = cv2.minAreaRect(contours[0])
box = cv2.boxPoints(rect)
box = np.int0(box)
global activeImage
activeImage = cv2.drawContours(origImage.copy(), [box], 0, (0, 0, 255), 4)
self.updateActiveImg()
def findMinCircle(self):
contours = self.findContours()
(x, y), radius = cv2.minEnclosingCircle(contours[0])
center = (int(x), int(y))
radius = int(radius)
global activeImage
activeImage = cv2.circle(origImage.copy(), center, radius, (0, 255, 0), 4)
self.updateActiveImg()
def fitEllipse(self):
contours = self.findContours()
ellipse = cv2.fitEllipse(contours[0])
global activeImage
activeImage = cv2.ellipse(origImage.copy(), ellipse, (0, 255, 0), 4)
self.updateActiveImg()
def matchTemplate(self):
# sets method equal to what is in the Template Method combo box
method = 0
if self.ui.cbTemplateMethod.currentText() == "SQDIFF":
method = cv2.TM_SQDIFF
elif self.ui.cbTemplateMethod.currentText() == "SQDIFF_NORMED":
method = cv2.TM_SQDIFF_NORMED
elif self.ui.cbTemplateMethod.currentText() == "CCORR":
method = cv2.TM_CCORR
elif self.ui.cbTemplateMethod.currentText() == "CCORR_NORMED":
method = cv2.TM_CCORR_NORMED
elif self.ui.cbTemplateMethod.currentText() == "CCOEFF":
method = cv2.TM_CCOEFF
elif self.ui.cbTemplateMethod.currentText() == "CCOEFF_NORMED":
method = cv2.TM_CCOEFF_NORMED
# perform the cv2.matchTemplate function and store output to 'result'
result = cv2.matchTemplate(workingImage, templateImage, method)
# find the minimum and maximum match values as well as their locations on the photo
min_val, max_val, min_loc, max_loc = cv2.minMaxLoc(result)
# If the method is TM_SQDIFF or TM_SQDIFF_NORMED, take minimum
if method in [cv2.TM_SQDIFF, cv2.TM_SQDIFF_NORMED]:
top_left = min_loc
else: # all other methods
top_left = max_loc
# height and width of template photo
h, w = templateImage.shape[:2]
# draw a rectangle on the active image
bottom_right = (top_left[0] + w, top_left[1] +h)
global activeImage
# create a copy of the workingImage object since cv2.rectangle modifies the object directly
# omitting the following line causes cv2.rectangle draw a rectangle on all references to the image object
activeImage = workingImage.copy()
cv2.rectangle(activeImage, top_left, bottom_right, 255, 4)
self.updateActiveImg()
def templateSURF(self):
if 'surf' not in globals():
global surf
print("creating surf component")
surf = cv2.xfeatures2d.SURF_create(400)
# find the keypoints and descriptors with SURF
kpT, desT = surf.detectAndCompute(templateImage, None)
kpA, desA = surf.detectAndCompute(activeImage, None)
FLANN_INDEX_KDTREE = 0
index_params = dict(algorithm=FLANN_INDEX_KDTREE, trees=5)
search_params = dict(checks=50)
flann = cv2.FlannBasedMatcher(index_params, search_params)
matches = flann.knnMatch(desT, desA, k=2)
# store all the good matches as per Lowe's ratio test
good = []
for m, n in matches:
if m.distance < 0.7*n.distance:
good.append(m)
MIN_MATCH_COUNT = self.ui.sbFeatureMinCnt.value()
print(len(good))
if len(good)>MIN_MATCH_COUNT:
src_pts = np.float32([ kpT[m.queryIdx].pt for m in good ]).reshape(-1,1,2)
dst_pts = np.float32([ kpA[m.trainIdx].pt for m in good ]).reshape(-1,1,2)
M, mask = cv2.findHomography(src_pts, dst_pts, cv2.RANSAC,5.0)
matchesMask = mask.ravel().tolist()
h, w = activeImage.shape
pts = np.float32([ [0, 0], [0,h-1], [w-1, h-1], [w-1, 0] ]).reshape(-1,1,2)
dst = cv2.perspectiveTransform(pts, M)
global activeImage
activeImage = activeImage.copy()
activeImage = cv2.polylines(activeImage, [np.int32(dst)], True, 255, 3, cv2.LINE_AA)
else:
print "Not enough matches are found -%d/%d" % (len(good), MIN_MATCH_COUNT)
matchesMask = None
draw_params = dict(matchColor = (0,255, 0), # draw matches in green color
singlePointColor = None,
matchesMask = matchesMask, # draw only inliers
flags = 2)
matchedImage = cv2.drawMatches(templateImage, kpT, activeImage, kpA, good, None, **draw_params)
cv2.imshow("matched image", matchedImage)
def picamSnap(self):
# initialize the camera and grab a reference to the raw camera capture
# set camera resolution to what is selected in the combo box
camera.resolution = map(int, str(self.ui.cbPicamResolution.currentText()).split("x"))
rawCapture = PiRGBArray(camera)
# allow the camera to warmup
time.sleep(0.1)
# grab an image from the camera
camera.capture(rawCapture, format="bgr")
image = rawCapture.array
global activeImage
activeImage = image
global workingImage
workingImage = image
global origImage
origImage = image
self.updateActiveImg()
self.updateWorkingImg()
self.updateOrigImg()