Sign upAdd Your SDK

Find objects in photos and video.

The Mobile Vision API provides a framework for finding objects in photos and video. The framework includes detectors, which locate and describe visual objects in images or video frames, and an event driven API that tracks the position of those objects in video.

Currently, the Mobile Vision API includes face and barcode detectors, which can be applied separately or together.

Main Features

Face API - The Face API captures details about the people that can be seen through the camera on a smartphone or tablet. Developers can get information about how many faces are present in the scene, where they are positioned, and their orientation. Additional details are available for each face, including the position of their eyes, nose, cheeks, corners and bottom of the mouth, ears and ear tips. There are additional methods for determining if a subject's eyes are open and if it looks like they are smiling.

Barcode API - The Barcode API detects barcodes in real-time, on device, in any orientation. It can also detect multiple barcodes at once.

It detects the following barcode formats:

  • 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
It automatically parses QR Codes and PDF-417 values, for the following supported formats: URL,Contact, information (VCARD, etc.), Calendar event, Email, Phone, SMS, ISBN, WiFi, Geo-location (latitude and longitude), Driver license/ID car.

SDK for Android

These are issues that apply to all of the APIs (Faces and Barcodes).

  •  MultiProcessor, maxGapFrames parameter: If multiprocessor maxGapFrames is set to 0, then trackers are never finalized. This can lead to "ghost" items for tracked objects that used to exist but are no longer present. Workaround: Do not set maxGapFrames 0. The defect is not present with the default value for the parameter. 
  • MultiProcessor (enhancement): Add notification for the onMissing event to the associated trackers. For intermediate frames in which a tracked item is not detected, this notifies the tracker that the item is currently not detected. User code can take the appropriate steps to better handle this situation (e.g., an app which displays a graphic overlay over detected items can hide those graphics during the brief periods that the item is not detected). 
  • Face.getPosition(), Face.getLandmarks(): The x coordinates are shifted to the left by a few pixels. The effect is more noticeable for smaller images. The shift will be corrected in a future release.

Required Permissions

  •  Internet
  •  Camera


Example A

Similar SDKs



Rate Now

Start a new discussion about Mobile Vision


More from this Developer