The Gini Vision Library provides Activities and Fragments for capturing, reviewing and analyzing photos of invoices and remittance slips.
By integrating this library into your application you can allow your users to easily take pictures of documents, review them and - by implementing the necessary networking interfaces - upload the document to the Gini API for analysis.
Communication with the Gini API is not part of the Gini Vision Library in order to allow clients the freedom to use a networking implementation of their own choosing. The quickest way to add networking is to use the Gini Vision Network Library. You may also use the Gini API SDK for Android or implement communication with the Gini API yourself.
The Gini Vision Library can be integrated in two ways, either by using the Screen API or the Component API. The Screen API provides Activities for easy integration that can be customized in a limited way. The screen and configuration design is based on our long-lasting experience with integration in customer apps. In the Component API we provide Fragments for advanced integration with more freedom for customization. We strongly recommend keeping in mind our UI/UX guidelines, however.
Customization of the Views is provided mostly via overriding of app resources: dimensions, strings, colors, texts, etc. Onboarding can also be customized to show your own pages, each consisting of an image and a short text.
The Gini Vision Library can be used on smartphones and tablets, too. On smartphones it has been designed for portrait orientation only and will always switch to portrait orientation in both Screen API and Component API usage. On tablets both portrait and landscape orientations are supported.
It is not required to limit your Activities extending the Screen API's abstract Activities or your Activities hosting the Component API's Fragments to portrait orientation. The Gini Vision Library takes care of limiting to portrait on smartphones.
Due to in-memory image handling applications using the Gini Vision Library must enable large heap.
We enabled landscape orientation and adapted some UI elements to offer a better experience to tablet users. We also removed the camera flash requirement for tablets since many tablets with at least 8MP cameras don't have an LED flash (like the popular Samsung Galaxy Tab S2).
For more information please consult our guide for supporting tablets.
Note: Please see our minimum hardware recommendations for tablets below. We recommend implementing corresponding hardware checks for the Gini Vision Library to deliver optimal results to users. As mentioned many tablets with at least 8MP cameras don't have an LED flash (like the popular Samsung Galaxy Tab S2) and we don't require flash for tablets. For this reason the extraction quality on those tablets might be lower compared to smartphones.
Furhter documentation can be found in our
The Gini Vision Library consists of four main screens:
- Onboarding: Provides useful hints to the user on how to take a perfect photo of a document.
- Camera: The actual camera screen to capture the image of the document.
- Review: Offers the opportunity to the user to check the sharpness of the image and to rotate it into reading direction, if necessary.
- Analysis: Provides a UI for the analysis process of the document by showing the user a loading indicator and the image of the document.
The Screen API provides a main Activity with which to start the Gini Vision Library. In previous
versions two abstract Activities which had to be overriden to react to events coming from the Gini
Vision Library. This has been changed and now only the implementations of the
GiniVisionNetworkService and GiniVisionNetworkApi have to be provided with the help of the
GiniVision class.
In order to support the widest variety of Android versions while keeping the look and feel consistent, we use the Android Support Library and provide only Activities subclassing the AppCompatActivity.
The Component API provides Fragments which you can include into your own layouts. This allows you more freedom to customize the Gini Vision Library, without being restricted to AppCompatActivities and the Gini Vision Library Theme.
To allow usage even if you aren't using the Android Support Library, we have been providing a Standard and a Compat version of Fragments in previous versions. We still provide them for backwards compatibility, but for new Fragments we will only provide versions using the Android Support Library.
We are providing example apps for the Screen API and the Component API. These apps demonstrate how to integrate the Gini Vision Library and how to use it with the Gini Vision Network Library to analyze photos of documents.
We also provide a separate standalone example app. This is more like a real world app and serves as an additional help for you to discover how the Gini Vision Library (and the Gini Vision Network Library) should be used.
Screen API: Android 4.4+ (API Level 19+) Component API: Android 4.4+ (API Level 19+)
- Back-facing camera with auto-focus and flash.
- Minimum 8MP camera resolution.
- Minimum 512MB RAM.
- Back-facing camera with auto-focus.
- Minimum 8MP camera resolution.
- Minimum 512MB RAM.
To install add our Maven repo to the root build.gradle file and add it as a dependency to your app module's build.gradle.
build.gradle:
repositories {
maven {
url 'https://repo.gini.net/nexus/content/repositories/open
}
}
app/build.gradle:
dependencies {
implementation 'net.gini:gini-vision-lib:4.3.1'
}
Gini Vision Library and the Gini Vision Network Library are available under a commercial license. See the LICENSE file for more info.