Vision SDK | Android Docs

The Mapbox Vision SDK is a software library that allows you to use powerful neural networks to run on existing devices to understand the roadway in real time.

Main SDK features are:

AR navigation: Build navigation with global coverage with turn-by-turn directions and custom objects in augmented reality
Scene segmentation: Expose clearly distinguished cars, lanes, curbs, and other road surfaces like sidewalks
Sign detection: Recognize road signs for speed limits, construction, turn restrictions, and more (see all sign types)
Safety alerts: Create custom alerts for speeding, tailgating, and potential collisions
Object detection: Detect and track nearby vehicles, bicycles, pedestrians, road signs, construction cones, and traffic lights
Lane detection: Detect the vehicle’s current lane to enable lane-level navigation
External camera support: Use custom integration camera implementations

SDK structure

The Vision SDK is composed of three frameworks you can interact with: Vision, Vision AR, and Vision Safety. Both Vision AR and Vision Safety depend on the Vision framework.

The Vision framework is the primary library, needed for any application that integrates Mapbox Vision. Its components enable camera configuration, display of classification, detection, and segmentation layers, lane feature extraction, and other interfaces. The Vision SDK's segmentation provides developers with the following pieces of lane information: number of lanes, lane widths, lane edge types, and directions of travel for each lane. A set of points describing each lane edge is also available.
The Vision AR framework is an add-on module for Vision used to create customizable augmented reality experiences. It allows configuration of the user’s route visualization: AR lane, AR fence, and their materials (shaders, textures), geometry and occlusion. Read more in the AR guide.
The Vision Safety framework is an add-on module for Vision used to create customizable alerts for speeding, nearby vehicles, cyclists, pedestrians, lane departures, and more. It notifies drivers about road conditions and potential hazards. For example, developers can track speed limits and other critical signage using sign classification and track the most recently observed speed limit. Program alerts to be triggered when the detected speed of the vehicle is more than the last observed speed limit. Read more in the Safety alerts guide.

Requirements

Vision SDK supports architectures arm64-v8a and armeabi-v7a and requires Android 6 (API 23) or higher. Some of devices that will work with the Vision SDK:

Samsung Galaxy S8, S8+, S9, S9+, S10, S10+, Note 8 and newer
Huawei P10, Mate 10 and newer
Xiaomi Mi 6 and newer
HTC U11 and newer
OnePlus 5 and newer
Other devices with Qualcomm Adreno GPU (Snapdragon 650, 7xx, 8xx series), Mali GPU (Huawei Kirin, Samsung Exynos) or PowerVR GPU. OpenCL support is required. Devices that do not support OpenCL, including the Google Pixel series, will not work with the Vision SDK.

To work with only the foundational components of Vision (segmentation, detection, and classification layers), developers need only import the Vision SDK.

Use of the Vision SDK in production requires that the device is pointed with a view of the road. We strongly recommend using a dashboard or windshield mount to keep your phone oriented correctly while you drive.

We have tested a few options and have seen positive results with two mounts (option 1 and option 2).

During the development you can set up the testing environment. Find more details at Testing and Development section.

Attribution

While the Vision SDK is using the camera you must display the Mapbox watermark on screen. Read more about attribution requirements in our terms of service.

Additional Developer Resources

Mapbox Developer Discord

Developer Cheatsheet

Mapbox Support

Ask AI

Was this page helpful?