Related
I've download the opencv from https://opencv.org/opencv-demonstrator-gui/ to make some live test on some images.
I found that this filter work perfectly for my needs:
,
I need to code it in my python script, tried to follow this tutorial :https://docs.opencv.org/3.4/d2/d2c/tutorial_sobel_derivatives.html
but I'm unable to find and match setting I need (pre-filtering Deriche, or Schar operator type).
I guess also I should use this syntax:
cv.Sobel(gray, ddepth, 1, 0, ksize=3, scale=scale, delta=delta, borderType=cv.BORDER_DEFAULT)
Thx.
UPDATE
Using this lines I'm close to right result:
scale = 1
delta = 0
ddepth = cv2.CV_16S
src = cv2.imread(image, cv2.IMREAD_COLOR)
src = cv2.GaussianBlur(src, (3, 3), 0)
gray = cv2.cvtColor(src, cv2.COLOR_BGR2GRAY)
grad_x = cv2.Sobel(gray, ddepth, 1, 0, ksize=3, scale=scale, delta=delta, borderType=cv2.BORDER_DEFAULT)
# Gradient-Y
# grad_y = cv.Scharr(gray,ddepth,0,1)
grad_y = cv2.Sobel(gray, ddepth, 0, 1, ksize=3, scale=scale, delta=delta, borderType=cv2.BORDER_DEFAULT)
abs_grad_x = cv2.convertScaleAbs(grad_x)
abs_grad_y = cv2.convertScaleAbs(grad_y)
grad = cv2.addWeighted(abs_grad_x, 0.5, abs_grad_y, 0.5, 0)
You are only doing the X derivative Sobel filter in Python/OpenCV. It is likely you really want the gradient magnitude, not the X directional derivative. To compute the magnitude, you need both the X and Y derivatives and then compute the magnitude. You also like will need to compute as float so as not to get one sided derivatives. You can later convert the magnitude to 8-bit if you want.
gradx = cv2.Sobel(gray,cv2.CV_64F,1,0,ksize=3)
grady = cv2.Sobel(gray,cv2.CV_64F,0,1,ksize=3)
gradmag = cv2.magnitude(gradx,grady)
The Scharr is similar and can be found at https://docs.opencv.org/master/d4/d86/group__imgproc__filter.html#gaa13106761eedf14798f37aa2d60404c9
I am trying to detect logo in invoices. Though I am able to get some results but not sufficient enough to process. While detecting logos, Unwanted text is also getting detected.
The following is from actual invoice:-original Image
and the following results I am getting Image after operations
I am using the`following code which I have written:-
gray=cv2.imread("Image",0)
ret,thresh1 = cv2.threshold(gray,180,255,cv2.THRESH_BINARY)
kernel_logo = np.ones((10,10),np.uint8)
closing_logo = cv2.morphologyEx(thresh1,cv2.MORPH_CLOSE,kernel_logo,
iterations = 1)
n=3
noise_removed_logo = cv2.medianBlur(closing_logo, n)
eroded_logo = cv2.erode(noise_removed_logo,kernel_logo, iterations = 8)
dilated_logo=cv2.dilate(eroded_logo,kernel_logo, iterations=3)
Could you please help me what changes should I make to remove noise from my documented image. I am new to Computer Vision
Few more sample:- Original document
The result I am getting:- Result after operations on document
Hello Mohd Anas Khan .
Your approch to define logo is too simple so it couldn't work. If you want a product-level approach, use some machine learning or deep learning. If you want just some toys, then a simple countours finder with fixed rules should work.
For example, in the following approach i defined "logo" as "the contour which has biggest area". You'll need more rules later, so good luck.
import numpy as np
import cv2
im = cv2.imread('contours_1.jpg')
imgray = cv2.cvtColor(im,cv2.COLOR_BGR2GRAY)
ret,thresh = cv2.threshold(imgray,127,255, cv2.THRESH_BINARY_INV)
rect_kernel = cv2.getStructuringElement(cv2.MORPH_RECT, (5, 5))
threshed = cv2.morphologyEx(thresh, cv2.MORPH_CLOSE, rect_kernel)
cv2.imwrite("contours_1_thres.jpg", threshed)
im2, contours, hierarchy = cv2.findContours(threshed,cv2.RETR_TREE,cv2.CHAIN_APPROX_SIMPLE)
ws = []
hs = []
areas = []
for contour in contours:
area = cv2.contourArea(contour)
x, y, w, h = cv2.boundingRect(contour)
print("w: {}, h: {}, area: {}".format(w, h, area))
ws.append(w)
hs.append(h)
areas.append(area)
max_idx = np.argmax(areas)
cv2.drawContours(im, [contours[max_idx]], -1, (0, 255, 0), 3)
# cv2.drawContours(im, contours, -1, (0, 255, 0), 3)
cv2.imwrite("contours_1_test.jpg", im)
The output images are as follow : (The detected logo is covered in green box )
I’m having issues with the following code (I’ve cut out large pieces but I can add them back in – these seemed like the important parts). In my main code, I set up a plot (“sectionizePlot”) which is a simple variation on another whiskered-plot
I’m looking to update them on the fly. In the same script, I’m using a heatmap (“ModifiedGenericHeatMap”) which updates fine.
Any ideas how I might update my whiskered-plot? Updating the ColumnDataSource doesn’t seem to work (which makes sense). I’m guessing that I am running into issues with adding each circle/point individually onto the plot.
One idea would be to clear the plot each time and manually add the points onto the plot, but it would need to be cleared each time, which I’m unsure of how to do.
Any help would be appreciated. I’m just a lowly Scientist trying to utilize Bokeh in Pharma research.
def ModifiedgenericHeatMap(source, maxPct):
colors = ["#75968f", "#a5bab7", "#c9d9d3", "#e2e2e2", "#dfccce", "#ddb7b1", "#cc7878", "#933b41", "#550b1d"]
#mapper = LinearColorMapper(palette=colors, low=0, high=data['count'].max())
mapper = LinearColorMapper(palette=colors, low=0, high=maxPct)
TOOLS = "hover,save,pan,box_zoom,reset,wheel_zoom"
globalDist = figure(title="derp",
x_range=cols, y_range=list(reversed(rows)),
x_axis_location="above", plot_width=1000, plot_height=400,
tools=TOOLS, toolbar_location='below')
globalDist.grid.grid_line_color = None
globalDist.axis.axis_line_color = None
globalDist.axis.major_tick_line_color = None
globalDist.axis.major_label_text_font_size = "5pt"
globalDist.axis.major_label_standoff = 0
globalDist.xaxis.major_label_orientation = pi / 3
globalDist.rect(x="cols", y="rows", width=1, height=1,
source=source,
fill_color={'field': 'count', 'transform': mapper},
line_color=None)
color_bar = ColorBar(color_mapper=mapper, major_label_text_font_size="5pt",
ticker=BasicTicker(desired_num_ticks=len(colors)),
# fix this via using a formatter with accounts for
formatter=PrintfTickFormatter(format="%d%%"),
label_standoff=6, border_line_color=None, location=(0, 0))
text_props = {"source": source, "text_align": "left", "text_baseline": "middle"}
x = dodge("cols", -0.4, range=globalDist.x_range)
r = globalDist.text(x=x, y=dodge("rows", 0.3, range=globalDist.y_range), text="count", **text_props)
r.glyph.text_font_size = "8pt"
globalDist.add_layout(color_bar, 'right')
globalDist.select_one(HoverTool).tooltips = [
('Well:', '#rows #cols'),
('Count:', '#count'),
]
return globalDist
def sectionizePlot(source, source_error, type, base):
print("sectionize plot created with typ: " + type)
colors = []
for x in range(0, len(base)):
colors.append(getRandomColor())
title = type + "-wise Intensity Distribution"
p = figure(plot_width=600, plot_height=300, title=title)
p.add_layout(
Whisker(source=source_error, base="base", upper="upper", lower="lower"))
for i, sec in enumerate(source.data['base']):
p.circle(x=source_error.data["base"][i], y=sec, color=colors[i])
p.xaxis.axis_label = type
p.yaxis.axis_label = "Intensity"
if (type.split()[-1] == "Row"):
print("hit a row")
conv = dict(enumerate(list("nABCDEFGHIJKLMNOP")))
conv.pop(0)
p.xaxis.major_label_overrides = conv
p.xaxis.ticker = SingleIntervalTicker(interval=1)
return p
famData = dict()
e1FractSource = ColumnDataSource(dict(count=[], cols=[], rows=[], index=[]))
e1Fract = ModifiedgenericHeatMap(e1FractSource, 100)
rowSectTotSource = ColumnDataSource(data=dict(base=[]))
rowSectTotSource_error = ColumnDataSource(data=dict(base=[], lower=[], upper=[]))
rowSectPlot_tot = sectionizePlot(rowSectTotSource,rowSectTotSource_error, "eSum Row", rowBase)
def update(selected=None):
global famData
famData = getFAMData(file_source_dt1, True)
global e1Stack
e1Fract = (famData['e1Sub'] / famData['eSum']) * 100
e1Stack = e1Fract.stack(dropna=False).reset_index()
e1Stack.columns = ["rows", "cols", "count"]
e1Stack['count'] = e1Stack['count'].apply(lambda x: round(x, 1))
e1FractSource.data = dict(cols=e1Stack["cols"], count=(e1Stack["count"]),
rows=e1Stack["rows"], index=e1Stack.index.values, codon=wells, )
rowData, colData = sectionize(famData['eSum'], rows, cols)
rowData_lower, rowData_upper = getLowerUpper(rowData)
rowBase = list(range(1, 17))
rowSectTotSource_error.data = dict(base=rowBase, lower=rowData_lower, upper=rowData_upper, )
rowSectTotSource.data = dict(base=rowData)
rowSectPlot_tot.title.text = "plot changed in update"
layout = column(e1FractSource, rowSectPlot_tot)
update()
curdoc().add_root(layout)
curdoc().title = "Specs"
print("ok")
Curious how one might create a plot with only text information. This will essentially be a "print" for the plot window.
The best option I've found so far is the following:
library(RGraphics)
library(gridExtra)
text = paste("\n The following is text that'll appear in a plot window.\n",
" As you can see, it's in the plot window",
" One might imagine useful informaiton here")
grid.arrange(splitTextGrob(text))
However, one doesn't have control (as far as I can tell) over font type, size, justification and so on.
You can do this using base graphics. First you'll want to take away all of the margins from the plot window:
par(mar = c(0,0,0,0))
And then you'll plot an empty plot:
plot(c(0, 1), c(0, 1), ann = F, bty = 'n', type = 'n', xaxt = 'n', yaxt = 'n')
Here's a guide to what's going on here (use ?plot.default and ?par for more details):
ann - Display Annotoations (set to FALSE)
bty - Border Type (none)
type - Plot Type (one that produces no points or lines)
xaxt - x axis type (none)
yaxt - y axis type (none)
Now to plot the text. I took out the extra spaces because they didn't seem to be necessary.
text(x = 0.5, y = 0.5, paste("The following is text that'll appear in a plot window.\n",
"As you can see, it's in the plot window\n",
"One might imagine useful informaiton here"),
cex = 1.6, col = "black")
Now to restore the default settings
par(mar = c(5, 4, 4, 2) + 0.1)
I hope that helps!
You could use annotate in ggplot2 like
library(ggplot2)
text = paste("\n The following is text that'll appear in a plot window.\n",
" As you can see, it's in the plot window\n",
" One might imagine useful information here")
ggplot() +
annotate("text", x = 4, y = 25, size=8, label = text) +
theme_void()
And you can of course remove the plot margins, axes, etc. to have just the text
Here's a handy example to play with too:
par(mar = c(0,0,0,0))
plot(c(0, 1), c(0, 1), ann = F, bty = 'n', type = 'n', xaxt = 'n', yaxt = 'n')
text(x = 0.34, y = 0.9, paste("This is a plot without a plot."),
cex = 1.5, col = "black", family="serif", font=2, adj=0.5)
text(x = 0.34, y = 0.6, paste(" Perhpas you'll:"),
cex = 1.2, col = "gray30", family="sans", font=1, adj=1)
text(x = 0.35, y = 0.6, paste("Find it helpful"),
cex = 1.2, col = "black", family="mono", font=3, adj=0)
Read up on ?par . There is limited capability to select the font type via the family and font arguments.
I have some data,
calvarbyruno.1<-structure(list(Nominal = c(1, 3, 6, 10, 30, 50, 150, 250), Run = structure(c(1L,
1L, 1L, 1L, 1L, 1L, 1L, 1L), .Label = c("1", "2", "3"), class = "factor"),
PAR = c(1.25000000000000e-05, 0.000960333333333333, 0.00205833333333334,
0.00423333333333333, 0.0322333333333334, 0.614433333333334,
1.24333333333333, 1.86333333333333), PredLin = c(-0.0119152187070942,
0.00375925114245899, 0.0272709559167888, 0.0586198956158952,
0.215364594111427, 0.372109292606959, 1.15583278508462, 1.93955627756228
), PredQuad = c(-0.0615895732702735, -0.0501563307416599,
-0.0330831368244257, -0.0104619953693943, 0.100190275883806,
0.20675348710041, 0.6782336426345, 1.04748729725370)), .Names = c("Nominal",
"Run", "PAR", "PredLin", "PredQuad"), row.names = c(NA, 8L), class = "data.frame")
calweight <- -2
for which I've created both a linear and a quadratic lm model
callin.1<-lm(PAR~Nominal,data=calvarbyruno.1,weight=Nominal^calweight)
calquad.1<-lm(PAR~Nominal+I(Nominal^2),data=calvarbyruno.1,weight=Nominal^calweight)
I can then plot my data values using ggplot2
qplot(PAR,Nominal,data=calvarbyruno.1)
But can't work out how to overlay a line representing the two lm objects... Any ideas ?
The easiest option is to use geom_smooth() and let ggplot2 fit the model for you.
ggplot(calvarbyruno.1, aes(y = PAR, x = Nominal, weight=Nominal^calweight)) +
geom_smooth(method = "lm") +
geom_smooth(method = "lm", formula = y ~ poly(x, 2), colour = "red") +
geom_point() +
coord_flip()
Or you can create a new dataset with the predicted values.
newdata <- data.frame(Nominal = pretty(calvarbyruno.1$Nominal, 100))
newdata$Linear <- predict(callin.1, newdata = newdata)
newdata$Quadratic <- predict(calquad.1, newdata = newdata)
require(reshape2)
newdata <- melt(newdata, id.vars = "Nominal", variable.name = "Model")
ggplot(calvarbyruno.1, aes(x = PAR, y = Nominal, weight=Nominal^calweight)) +
geom_line(data = newdata, aes(x = value, colour = Model)) +
geom_point()
Earlier I asked a related question and Hadley had this good answer. Using the predict function from that post you can add two columns to your data. One for each model:
calvarbyruno.1$calQuad <- predict(calquad.1)
calvarbyruno.1$callin <- predict(callin.1)
Then it's a matter of plotting the point and adding each model in as a line:
ggplot() +
geom_point(data=calvarbyruno.1, aes(PAR, Nominal), colour="green") +
geom_line(data=calvarbyruno.1, aes(calQuad, Nominal), colour="red" ) +
geom_line(data=calvarbyruno.1, aes(callin, Nominal), colour="blue" ) +
opts(aspect.ratio = 1)
And that results in this nice picture (yeah the colors could use some work):
(source: cerebralmastication.com)