I am using OpenCV on iOS to detect a rectangular label to assist users in snapping a photo of that label. I have an overlay that presents once the matches threshold is met.
My question is, does that patch image used have to be exact? The labels I am detecting have text on them that vary from label to label. All the same font but different characters. Is it possible to train OpenCV with a patch images color and/or size/dimensions? Or is there perhaps another way around this issue?
Here is a close example to the labels Im scanning, EXCEPT THERE ARE NO ICONS AND ALL ONE FONT TYPE.
Here is the tutorial I am following, which is achieved with an image of a target. http://www.raywenderlich.com/59999/make-augmented-reality-target-shooter-game-opencv-part-3
Related
I am using QSplashScreen to display a splash screen for my QtQuick 2 application. I am using a decent quality image (838 X 500). It displays correctly in some screen resolutions. But for some other resolutions like 2K resolution, it looks like some low quality transformation is applied.
Already tried Qt::SmoothTransformation
Connecting my laptop to my monitor (I have 2K monitor) and selecting 'Duplicate display' option in display management, produces low quality image. Selecting 'Second screen only' produces decent quality image.
QPixmap pixmap(qApp->applicationDirPath()+"/splash.png");
QSplashScreen splash(pixmap.scaledToWidth(screen_width*0.35,Qt::SmoothTransformation));
splash.show();
Is there a way to display an image without loosing the quality?
I am experiencing the same issue with 'Image' in QML.
If Qt is attempting to render the image at the same "physical" size on a higher resolution display, then it will have no choice but to upscale the image. You should provide a higher resolution image:
https://doc.qt.io/qt-5/scalability.html#loading-files-depending-on-platform
The target platforms might automate the loading of alternative
resources for different display densities in various ways. On iOS, the
#2x filename suffix is used to indicate high DPI versions of images.
The Image QML type and the QIcon class automatically load #2x versions
of images and icons if they are provided. The QImage and QPixmap
classes automatically set the devicePixelRatio of #2x versions of
images to 2, but you need to add code to actually use the #2x
versions:
if ( QGuiApplication::primaryScreen()->devicePixelRatio() >= 2 ) {
imageVariant = "#2x";
} else {
imageVariant = "";
}
Forget the part about iOS - I think it's outdated, because it has applied to all platforms for a while now.
So in your case, you should add a splash#2x.png that is twice the resolution (and detail) of splash.png. If you're just using QPixmap, you will need to add the code above to ensure the correct image variant is selected. If you're using QML's Image type, it will select it automatically.
I have small application in famo.us framework. There are 5 images. I want drag images. requirement is like when i drag firstImage, second image should be visible behind the first image. i tried to show second image on dragging up of firstImage, but it hides firstimage.
so is it possible to show second image behind the first image ??
Thanks
Check the z position of your images. Chances are you're encountering z-fighting, where two elements exist on the same position on the same z-axis. Try changing the z-index of the images to different values first to see if that fixes it.
Otherwise, check the 3d-transform you've supplied to famous to make sure your images aren't resting on the same z-value.
I want to know if exist templates to add text on photoshop like in the following pictures:
http://sphotos-e.ak.fbcdn.net/hphotos-ak-ash4/485070_566956359990277_263279122_n.jpg
https://fbcdn-sphotos-b-a.akamaihd.net/hphotos-ak-ash4/3263_403189956409783_609225582_n.jpg
Could you suggest some pages where I can study and try those effects ?
Thanks.
Step1: Find the fonts. Search on font sites for similiar fonts or try to cut out the font and upload it to http://www.whatfontis.com/ or http://www.myfonts.com/WhatTheFont/
Step2: just play around with blending options and stuff.
from what I can see it's these settings you want too look at:
"Notche Latina": Gradient overlay, white stroke and black dropshadow.
"photo recoblablabla": inner shadow and might be in overlay-mode or something
"Festival de la bira": once again, white stroke
First make a layer adjust the objcet on layer as per your requirement.
Add some effects from filter gallery if you required (or layer style).
Click on the text toolbar type your text (required text).Then to apply effects
right click on text layer and click on rasterize type.now you can add effects from the filter gallery to the text.(or Layer style eg-inner shadows,bevel and emboss).
I have an windows app which contains some dialogs. the dialogs have been built using mfc. I am drawing some images (.png) on every dialog using CImage::Draw() method. I want to mention that I am not using any picture contol on the dialog to render these images instead I am loading them at runtime using some handle.till this everything is ok. now when the image is loaded the background of those images are coming as white. the images in the resource file does not have the white background. my question is how to change the background of these images while drawing them on the dialog? I want the background of the image similar to the color of default dialog which i am using.
One more question the .png images are not rendering well(the images are scattered) in the dialogs of windows server 2008 R2 machine. what could be the possible remedy for this?
any help will be appreciated.
Your PNG images are obviously not 32-bit. You need an alpha channel and a transparent background. Open your images in e.g. Paint.NET. I bet your background is white there too! Regarding the image quality, are you stretching your images on draw?
Edit: For 8-bit imagers, I believe a call to SetTransparentColor is required. For 32-bit images, perhaps this function will do: TransparentBlt
I am building an OpenGL application. I read through the GLUI tutorial on Code Project to create windows form controls on an OpenGL Application. But my requirement is to develop a color choser/picker, like an RGB chart or RGB cube to select a color. The tutorial on Code project shows the list of colors as a drop down box. However that wont really help me, as I require it to be present as a windows color picker. I know that color picker as a dialog box is a part of the windows application. Can anyone suggest me a way to use it with my OpenGL Application?
You can try fox-toolkit. It is a C++ based Toolkit for developing Graphical User Interfaces. It provides Color Picker and OpenGL widgets for 3D graphical manipulations.
On windows, you can call directly ChooseColor() from the windows API. It will open the native color color chooser.If you need a cross-platform solution, tiny file dialogs on sourceforge also has a color picker and no main loop.
You could render a Quad/Triangle/Cirle with different color on each vertex and activate smooth shading for interpolation between these points. Then just read back the color value from OpenGL at the mouse position.
edit: or like that where you calculate the color on the mouse position by yourself (reading values slows down OpenGL a lot!): http://sharathpatali.wordpress.com/2009/07/07/a-color-picker-for-pymt/
I did this with simple gradient image which i kept in memory. Then i just tracked my mouse position on the image, and simply read the data from the image (that was kept in RAM) and get the 32bit RGBA color value for it. This is easier than reading the pixels from screen (also faster and more reliable).
This also allows a lot more flexible way of presenting the palette, only your imagination is the limit on the looks of your palette. Note: you must use 32bit colors on the image, because if you want smooth edges, you simply just fade the alpha but keep the colors the same, so the colors wont get distorted at edges. Don't forget to enable blending when rendering the image.