Image Analysis Toolset is an image analysis Swiss Knife that unlike others apps that use crowdsources tagging services, that involve people that manually add tags to pictures, it's entirely based on computer vision and deep learning. This means that only advanced artificial neural networks handle the loaded pictures without manual human interventions.
The feature set offered is extremely wide and includes:
► Element Identifier:
To identify elements of a picture and search info about them. It supports broad sets of categories ranging from inanimate objects to plants and animals.
► Web image detector:
To find info about the image, searching the Internet for similar images and related web pages, and guessing the content according the grabbed information. This feature provides you the related labels, links of the involved web pages, showing matching and visually similar images (if available), allowing you to save also the respective links or the image files.
► Optical Text Recognition (OCR):
To digitize the text of a picture or of a scanned document, so you can easily edit or put it wherever you want, or search info from its content.
► Logo Identifier:
To detect the logo of a product or a service and search related information.
► Landmark Identifier:
To detect popular natural and man-made structures within an image and search related information.
► Barcode Detector:
Can identify almost all types of barcodes. 1D barcodes: EAN-13, EAN-8, UPC-A, UPC-E, Code-39, Code-93, Code-128, ITF, Codabar; 2D barcodes: QR Code, Data Matrix, PDF-417, AZTEC.
► Face Insight:
Detect multiple faces within an image, along with the associated facial attributes and emotions. Compare faces to determine resemblance level. This feature set is also able to estimate the age range from facial features and identify celebrities.
With colorimeter you can identify all colors within an image and see their representation in RGB, HSB and HEX notation. For each detected color, the app will tell you the name of the color or the name of the most similar color, in case the color tone is uncommon and has no name.
► Censorship Risk Meter:
This tool allows you to check an image to determine if its content could be censored by automatic systems or lead to ban. This feature is useful since many social networks and websites perform automatic checks of the uploaded pictures and could take actions against the user if a critical content is detected.
To allow you to spot tampered sections in an image, according the inhomogeneity in the distribution of the error compared to the local pattern.
► EXIF Info:
This feature allows you to load and extract EXIF metadata from picture files, if available.
- You can enable the vocal output from IAT settings to configure the App to speak out loud results.
- Share a picture from whatever app with Image Analysis Toolset and Image Analysis Toolset will load your picture and when you select a feature, the selected picture will be analyzed directly.
- You can export the analysis results as text file.
- Element Identifier, Optical Text Recognition (OCR), Barcode Detector, Face Insight and EXIF Analysis can be used also without any Internet connection (however with an active internet connection, element identifier, text recognition and face insight are more accurate).
- Customizable Detection with self-trained models.
- Realtime Detection.
The App supports TalkBack and can be used also by low-vision users.
- CUSTOM MODELS UTILITY (Windows)
- CUSTOM MODELS UTILITY (Linux)
Under active development.