I have some images that their file name are corresponding GPS data (Long/Lat), and I'm trying to use preferably MATLAB to put them together and make a bigger image! In other words, I want to creat a local map by having several aerial pictures.
anyone has any idea or knows any software that really works to concatenate some images to make a bigger image?
Thanks
It depends on what type of images you have.
If you have georeferenced images, in GeoTiFF or other format, the Matlab Mapping Toolbox is very good for loading geographic imagery, manipulating it, and displaying it.
If you just have images that aren't georeferenced, you can use the Matlab Computer Vision Toolbox to register the images and stitch them together like a panorama. Photoshop also has some powerful panorama stitching tools if you don't need to do it programmatically.
Related
I am trying Tensorflow faster_rcnn_resnet101 model to detect multiple objects in a 642*481 image. I tried two ways but none got satisfactory results.
1) In this way, I cropped the objects (50*160 rectangle) from training images (training images might be at a different dim size than 642*481 test image). I use those cropped images to train the faster_rcnn_resnet101. Looks like it has good result if the test set is also a cropped image on the same size. But for the 642*481 test image, it could not detect multiple objects there with good results.
Then I think about maybe the model rescaled the test image to match the 50*160 so the details got lost. In this thought, I tried another way
2) I copied each cropped images into a 642*481 white background respectively (padding basically). So each training image has the same dim size as the test image. The location of the cropped images copied on the background is purposely set as random. However, it still does not have good detection on the test image.
To try its performance, I use GIMP to make windows that contain objects and replace other parts outside the windows with white pixels. The result is much better.
If we only keep one object window in the image and make other parts as white, the result is super.
So my question is what is happening behind those? How could I make it work to detect multiple objects in the test images successfully? Thanks.
The idea is that you should use training images containing multiple instances of objects during training. In other words, you don't have to crop ! Especially that your images are not very big.
However, you DO have to identify (label) the objects. This means to know the bounding box of each object in an image. Furthermore, this data should be assembled together with the original image in a tfrecord file. There is a good guide here with the whole process. In your case, you would label all objects in the source image (642x481), instead of just the "raccoon". If your objects have multiple classes, make sure you label them as so !
If you train with these images, which contain the objects in their contexts, then the network will learn to recognize similar images, which have objects in context.
Pretty much with every Pagespeed test I do for all my website I get the comment "Optimize images by lossless compressing image X" which often increases my page rank a lot.
I already save EVERY image with 'save for web' with Photoshop, but I was wondering how I could "Optimize images by compressing lossless" even more. As far as I know I'm already doing everything I can.
Really wondering..
Off-topic, but I noticed that Google's PageSpeed uses a Retina device to check, since all my Retina images got loaded instead of the regular ones. Since these are larger than the area I got a 1/100 score on the mobile segment. Haha.
This was a real issue with many of my sites, however I use the free version of kraken to 'loosely compress' all of my images and this passes the Google Test, thus boosting rankings!
https://kraken.io/web-interface
I must have used this for well over 10,000 images already!
The images you create in programs like Photoshop and Illustrator look amazing but often the file sizes are very large. This is because the images are made in a format that makes them easier to manipulate in different ways. If you put these files on your website it would be very slow to load. Optimizing your images for the web means saving or compiling your images in a web-friendly format depending on what the image contains.
How does it work?
There are two forms of compression that we need to understand, Lossy and Lossless.
Images saved in a lossy format will look slightly different than the original image when uncompressed. Keep in mind that this is only visible at a very close look. Lossy compression is good for web, because images use a small amount of memory, but can be sufficiently like the original image.
Images saved in lossless format retain all the information needed to produce the original image. For this reason, these images carry a lot more data and in return are a much large file size.
We also can optimize images for the web by saving them as the appropriate dimensions. Resizing the image on the webpage itself using CSS is helpful but the issue is the web browser will still download the entire original file, then resize it and display it.
Can you imagine taking a poster size image and using it as a thumbnail? The little 20px by 20px image would take as long to load as the original poster when we could just be loading a 20px image the whole time.
How to Optimize Images?
In simple terms optimizing your image works by removing all the unnecessary data that is saved within the image to reduce the file size of the image based on where it is being used in your website. Optimizing images for the web can reduce your total page load size by up to 80%.
Full optimization of images can be quite an art to perfect as there are such a wide variety of images you might be dealing with. Here are the most common ways to optimize your images for the web.
Reduce the white space around images – some developers use whitespace for padding which is a big no-no. Crop your images to remove any whitespace around the image and use CSS to provide padding.
Use proper file formats. If you have icons, bullets, or any graphics that don’t have too many colors use a format such as GIF and save the file with lower amounts of colors. If you have more detailed graphics then use JPG file format to save your images and reduce the quality.
Save your images in the proper dimensions. If you are having to use HTML or CSS to resize your images, stop right there. Save the image in the desired size to reduce the file size.
To resize your images you will have to use some form of program. For basic compression, you can use a simple editing program such as GIMP. For more advanced optimization you will have to save specific files in Photoshop, Illustrator, or Fireworks.
I want to implement an image processing method in order to distinguish scanner images from camera images. How can I collect images from scanners and cameras? The images should be very different in scale and content and capture devices and rgb images or gray-scale images.
thanks a lot.
The most famous database for machine learning is the UCI Machine Learning Repository, where you can find tons of open source datasets organized by learning objective, data type, number of instances and more.
I really want to learn how an image is composed (i.e. array of bits, or however, how is the color composed for each pixel, etc). Can you point me in the right direction? I'm not really sure what to search for.
Thanks a lot in advance.
So what I want to do is to be able to modify the picture pragmatically, i.e. change to black and white, scale it, crop it, etc, and for this I would really like to learn how the image is composed instead of just finding these algorithms online.
You don't always need to know low level mathematical details(matrixes,quantisation,fourier transform etc.) of graphic formats to manipulate images.
For all the things you want to do you may use proper libraries.
For example in PHP libraries used freuqently to manipulate images are:
GD - http://php.net/manual/en/book.image.php
ImageMagick - http://php.net/manual/en/book.imagick.php
It depends on the image format that you're interested in manipulating. Each format (more or less) is composed in a different manner, and based on that has a different set of capabilities for manipulating the image.
Different sets of actions on an image favor different image formats, as does the type of image you want to manipulate.
Provide more details about what you want to do with the image and I'm sure someone else will come along and tell you which formats are best and how they are handled.
What are some of the points that I need to follow if I want to have good quality images in a LaTeX document. These images are mostly screenshots of a software application or flow charts.
Below are two such images.
Flow Chart
Screenshot
Thanx
For diagrams, the rule is to use vector formats as much as you can — PDF, EPS or native LaTeX packages. When using vector graphics, the picture does not loose resolution and can be scaled freely. For a flow chart, I would either export it from the drawing application as a PDF, or use PGF/Tikz to produce it from LaTeX (see also examples). If your drawing application does not have a PDF export, consider using one that does — e.g., UMLet.
If you can't use vector graphics (e.g., because it is a screenshot), make sure you use high-enough resolution to begin with. If it is an academic paper, the publisher usually has guidelines for this.
If you use PDFLatex you can use png images and in those cases you definately should use png over jpeg. PNG compression is not lossy, so you get the best quality at the expense of file size.
The second important point is to create the images with sufficient resolution, for printing it should be about 300-600 dpi, higher is better but the filesize of the images and the resulting document will increase. For documents that will only be looked at a screen you can use a lower resolution, about 72-100 dpi should be enough.
For diagrams you should create vector graphics (eps or pdf) if possible, that way you do not lose any quality.
For screenshots, there is not much to do, but for flow charts, I'd suggest to create them in PDF format (vectorized) and to compile your LaTeX source with pdflatex.
for the flowchart i'd suggest TikZ, then your chart is directly typeset in TeX. Here's an example: http://www.texample.net/tikz/examples/simple-flow-chart/
Screenshots are pretty much a lost cause. I've had a good experience saving them as PDF and then embedding them, but you want to make sure you're on a high-res capture to begin with.
Charts are very easy. Most graphics programs (e.g., Vizio, OmniGraffle) will let you save it as EPS or PDF, and scaling works fairly well.