What is the role of Augmented Images Database?
I am using arcoreimg tool to generate the img database, but why do need to generate it. I know it is a very basic question please provide your guidance
I am new to AR following Google AR SDK, Please help.
The role of the AugmentedImageDatabase is to store a compressed representation of the images you'd like to track. This is used by ARCore to detect images in the real world.
Generating a database at compile time with the arcoreimg tool has several advantages:
Your app no longer needs to bundle the original PNG or JPEG files. Your app would instead include a smaller database file, resulting in a smaller APK size.
Your app no longer needs to decode the original PNG or JPEG files to extract the image's feature points at runtime. This is an operation which takes roughly ~30ms for each image.
Related
I'm looking for a way to inspect the the ICC color profile data provided CGColorSpace's copyICCData() method.
Specifically, I'm loading PNG images into UIImages on iOS, and trying to find a way to use the let iccData:CFData? = aUIImage.cgImage!.colorSpace!.copyICCData() to determine the gamma for the image file. This is for a game that uses 3D rendering— if the source image has a standard 2.2 gamma, I'll load the image data into a texture as sRGB (e.g. MTLPixelFormatRGBA8Unorm_sRGB) and if it has a gamma of 1.0 I'll instead load it as a linear texture (e.g. MTLPixelFormatRGBA8Unorm).
Note: The solution of just passing a UIImage/CGImage to the rendering system (SceneKit/Metal) and letting it sort it out won't work here because: 1. Some of the rendering I'm doing is assembling 2D images into a 3D texture, so that's something I need to do with raw data, not something I can just read from a standard image file format; 2. I'm specifically trying to pass gamma-1.0 images into the rendering system to avoid the overhead of sRGB→linear conversion (rendering is in linear space).
Also: Manual ICC-parsing solutions, Apple-API-using solutions, and open-source library suggestions are all acceptable answers. This is not specifically a query for tool recommendation — any solution that'll work is a good one — but in my research, manual ICC parsing would be unwieldy and Apple's APIs don't seem to expose any ICC properties. So I believe the most likely answer is a pointer to some library out there that I haven't been able to find via Google or GitHub or CocoaPods or StockOverflow, and will be gladly accepted.
Your best bet is to use sample icc 'https://sourceforge.net/projects/sampleicc/'. Just get the profile data as you described, then use OpenIccProfile to load it up. From there, get a ref to the header structure (.m_Header) and pull the info you need. I'd also recommend you take a look at RefIccMax 'https://github.com/InternationalColorConsortium/RefIccMAX' which is a newer version of the same lib, but not ready for primetime.
Is it possible to do very basic image recognition to compare an image against a database of images(resource folder images or any web servers images if we have) and determine which image in the database is the best match? I don't need to do any processing of any of the images, but simply differentiate between a finite list of images.
Is it any open source code available ?
I would recommend using OpenCV if you simply want to compare images (i.e. decide if two images are the same).
Here is a similar question on SO:
iOS image comparison
I would also go about reading a little bit about what Core Image (the iOS image library) has to offer, before going about OpenCV or other 3rd party.
I hope this helps.
I have a JPEG image stored in memory as a blob and am looking to apply some basic transformations to it (e.g. resize, convert to greyscale, rotate etc.)
I am currently using Google Scripts which doesn't have a native image library as far as I can tell.
Are there standard algorithms or similar which would allow me to work with the raw binary array, knowing it represents a JPEG image, to achieve such a transformation?
Not the answer you are looking for I guess, but...
To be able to do image processing using JPEG files as input, you need to decode the images. Well, actually, 90/180/270 degree rotation, flipping and cropping is possible as lossless operations, and thus without decoding the image data. But for anything more advanced, like resizing, you need to work with a decoded image.
Both the file structure (JIF/JFIF) and algorithms used to compress the image data in standard JPEG format are well defined and properly documented. But at the same time, the specification is quite complex. It's certainly doable if you have the time and know what you are doing. And if you are lucky, and your JPEG blobs are all written the same way, you might get away with implementing only some of the spec. But even then, you will need to (re-)implement large parts of the spec, and it might just not be worth it.
Using a 3rd party service to convert it for you, or create your own using a known library, like libjpeg or Java's ImageIO, etc. might be your best bets, if you need a quick solution, and don't have too strict requirements for performance.
There are no straightfoward image processing capabilities available in Apps Script. You'll have either expose your Python as a web service and call it from Apps Script or use the Drive REST API to access the files from your Python app or use any api webservices.
GAE Python has Image processing capabilities check the below url:
https://developers.google.com/appengine/docs/python/images/
Available image transformations
The Images service can resize, rotate, flip, and crop images, and enhance photographs. It can also composite multiple images into a single image.
I've asked similar questions before, but have not received a definitive answer. Seems that there must be a way to simply add/modify metadata to an image without loading the image into memory, without having to deal with directly reading bits.
Seems like ways exist when using CMSampleBufferRefs, but I need to be able to do this with a regular image already saved to disk.
For instance, given a very large png at /Documents/photo.png, I want to modify its exif metadata without having to load that image.
You can use libexif - I've had success with compiling it for iOS before. With libexif, you can modify any image's EXIF metadata.
If you know how to modify the EXIF, you can modify the binary data directly from the file. Just replace in the image the binary portion with the new one.
I don't know if objective-c permit this, but in ansi c should be simple. The complicate part is to identify the exact part to change.
I'm putting together a script to find remove duplicates in a large library of images. At the moment I'm doing a two pass filter of first finding files of the same size and then doing a sha256 on a 10240 byte piece of the file to get a fingerprint of the files with the same size (code here).
It works well, but I'm guessing there are probably checksums built in to the jpeg format that I could use instead of doing the sha256.
Does anyone know if there are checksums or other components that could act as checksums / fingerprints? If so, is there an efficient way to access them?
I don't think the JPEG specification includes any kind of checksum in the way you're describing.
A JPEG can contain a thumbnail as part of its EXIF metadata, though. It's not a perfect indicator, since it's possible for two different images to have the same thumbnail. There's at least one documented case of a thumbnail not being replaced after the image had undergone substantial modifications, said thumbnail revealing much more than the publisher had intended.
Its been awhile since I've dug into the IJG library, but I don't think there's an easy class member or function call you can use there to check for some type of fingerprint. You could use the built in EXIF tags if you can control the encoding of the images...
I'm just built a very similar script. I don't want to checksum metadata I want to see if the actual images are duplicates even if tags have been modified. Best for that is not to sort by size, but do sort by the checksum istelf. I use jhead to remove metadata and then checksum the whole file (but I also thought about just doing part of it, but actually I don't think it saves much time). jhead doesn't use shared memory (pipes) and does overwrite so I just copy the file to shared memory first. I place the checksum in the ImageDescription field for later faster retrieval. Obviously this also allows to check image integrity later and is part of why I checksum the whole thing. Tip: exiv2 is MUCH faster for reading and writing the metadata than exiftool for one at a time decision based manipulation.
In JPEG standard(ITU-T.81) i believe there isn't any field/syntax element which has a checksum or such, for the whole compressed jpeg image file. Unless a customised application puts such filed in the Application segment, or as meta data for which segments are provided in the standard.
So to serve your purpose, what you are doing is one soln.
Other could be some kind a application wrapper which will call some binary file compare utlitiy (like beyond compare, or even a windows command fc /b) and check the result of that compare utility and take the decision u want to.
-AD
One way you could perform is reduce all images to a fixed size and store that as a thumbnail. Then the image comparison would compare similar sized images and give you a chance of being a duplicate - useful if you have cropped (unless cropped heavily) or resized images and want to find those 'duplicates'.
In the XMP specification there are document ID and version ID which should uniquely identify the version of the image.
The problem with these (and with any other metadata-based identification method) is that it might not be respected by some applications that can change the content of the jpeg updating the metadata accordingly.