Mapillary - Detailed Review
Search Tools
Mapillary - Product Overview
Mapillary Overview
Mapillary is a crowdsourced street-level imagery platform that leverages computer vision to create and share detailed map data. Here’s a breakdown of its primary function, target audience, and key features:Primary Function
Mapillary’s main purpose is to collect, process, and make available street-level images from around the globe. These images are used to create a collaborative street-level view of the world, which helps in updating maps, extracting spatial data, and managing assets more effectively.Target Audience
The platform is open to a wide range of users, including hobbyists, researchers, commercial entities, cities, transport agencies, logistics and delivery companies, automotive companies, and humanitarian organizations. Anyone who needs geospatial information from street-level images can benefit from Mapillary.Key Features
Image Capture and Upload
Users can capture and upload street-level imagery using any camera, such as smartphones, GoPros, or dashcams. The platform provides tools like the Mapillary mobile app, Desktop Uploader, and command-line tools to make the upload process easy.Image Processing and Data Extraction
After uploading, the images are processed to blur faces and license plates. Mapillary’s computer vision technology then creates 3D reconstructions and recognizes objects in the images, including over 40 feature classes and 1,500 traffic sign classes.Map Updates and Asset Management
The extracted data can be used to update maps with details not visible from aerial images. This is particularly useful for asset management, as it allows organizations to capture and manage their assets more efficiently.Exploration and Sharing
Users can explore the world through smooth transitions between images using Mapillary’s web and mobile apps. The imagery and extracted data can be downloaded and shared on websites or social media.Integration and Open Data
Mapillary integrates seamlessly with OpenStreetMap and other GIS tools. The platform promotes open data, allowing everyone to access and use the imagery and extracted data to make better decisions about cities, roads, mobility, and navigation.Conclusion
In summary, Mapillary is a powerful tool for anyone looking to contribute to or utilize street-level imagery and map data, fostering a collaborative environment where everyone can benefit from shared data. Mapillary - User Interface and Experience
Mapillary Overview
Mapillary, a crowdsourced street-level imagery platform, offers a user-friendly and interactive interface that caters to various needs, from casual exploration to advanced data extraction.Search and Exploration
The Mapillary interface, particularly on the iOS app, is designed for ease of use. Users can search for any location using the search bar at the top of the map window. Once a location is selected, the map zooms in, allowing users to explore the available street-level imagery.Viewing Captures
To view captures, users simply zoom in and click on the desired image. This opens a split-screen view showing the imagery at the top and the map view at the bottom. The interface includes spatial arrows that allow users to move to the previous or next image in the current capture or step into a different nearby capture. Additional controls, such as playback buttons and a speed control icon (marked by a hare icon), enable users to adjust the viewing experience to their liking.Map Style and Customization
Users can toggle between different map styles, which also changes the basemap seen while capturing new imagery. This flexibility helps in customizing the visual experience according to the user’s preferences.Sharing and Reporting
The interface allows users to share images easily through a dedicated sharing option. There is also a feature to report images if they contain any issues or inaccuracies, ensuring the community can maintain the quality of the uploaded content.Data Extraction and Integration
For more advanced users, Mapillary provides tools to extract spatial data, which can be integrated into various applications, such as GIS tools, the Mapillary Python SDK, or the Mapillary API. This data can be used to update maps, manage assets, and support research projects, all of which are facilitated through a user-friendly and accessible interface.User Experience
The overall user experience is enhanced by the platform’s use of state-of-the-art Structure from Motion technology, which allows for smooth transitions between images. Despite some users reporting usability issues with recent updates to the mobile apps, the core functionality and user interface of Mapillary remain intuitive and accessible.Conclusion
In summary, Mapillary’s user interface is designed to be user-friendly, allowing individuals to easily capture, explore, and utilize street-level imagery. The platform’s tools and features are well-organized and straightforward, making it accessible for a wide range of users. Mapillary - Key Features and Functionality
Mapillary Overview
Mapillary is a sophisticated street-level imagery platform that leverages advanced AI and computer vision technologies to provide a range of powerful features for transportation agencies, urban planners, and other users. Here are the main features and functionalities of Mapillary, particularly focusing on its AI-driven capabilities:Street-Level Imagery Collection and Hosting
Mapillary allows users to upload and host street-level imagery, either as photologs of entire road networks or as spot collections using consumer-grade cameras and phones. This imagery is accessible via desktop and mobile applications, as well as through the Mapillary API and integrations with third-party geospatial tools like ArcGIS, Geocortex Essentials, and custom web maps.Automatic Feature Extraction
One of the key features of Mapillary is its ability to automatically extract various types of road features from the uploaded imagery using computer vision. This includes the detection of up to 96 types of road features, such as traffic signs, potholes, guardrails, and more. These features are extracted and made available in standard geospatial formats like shapefiles or GeoJSON.AI Detections and Semantic Segmentation
Mapillary’s AI technology processes imagery using semantic segmentation, which differentiates pixels into classes that denote different objects in the real world. This process segments images into up to 97 semantic classes, allowing for the identification of various objects within the images. This feature helps in quickly finding images where specific objects have been detected, which is particularly useful for map editing and other geospatial analyses.Point Features and Object Detection
The platform can detect and represent up to 42 types of objects as point features on the map, including items like trash cans and manholes. These detections provide estimated coordinates based on image analysis and can be further refined by triangulating the real-world location of objects detected in multiple images. Users can access these detections through the Mapillary API and web app.Map Editing and Integration
Mapillary’s AI detections are highly beneficial for map editing. Users can visualize and search for automatically detected features, such as zebra crossings, traffic signs, and other map elements. This feature is particularly useful when these elements are not visible in the images or are missing traffic signs. The detections can be integrated into tools like OpenStreetMap, making it easier to add or update map data.Customization and Developer Tools
Mapillary provides a range of tools for developers, including the MapillaryJS library, which is an interactive and extendable platform for street-level imagery and semantic mapping visualization. This library offers core capabilities and powerful extension APIs, allowing developers to customize and extend the street-level imagery experience with features like 3D models, animation, and custom data formats.Accessibility and Integration
The data and imagery stored on Mapillary are accessible across various platforms, including desktop and mobile applications. The platform also integrates with other geospatial tools, enabling users to review and edit data layers related to asset management, road surveys, crash analysis, safety assessment, and pavement maintenance.Conclusion
In summary, Mapillary’s AI-driven features make it a powerful tool for automatic feature extraction, map editing, and geospatial analysis, leveraging advanced computer vision and semantic segmentation to enhance the accuracy and efficiency of these processes. Mapillary - Performance and Accuracy
Performance and Accuracy
Mapillary’s approach to geospatial data collection is efficient and cost-effective, allowing communities to gather data faster and cheaper. Here are some key points on its performance and accuracy:Positional Accuracy
Studies have shown that Mapillary-generated data can achieve high positional accuracy. For example, in an experiment evaluating street light data, the overall positional accuracies were under 5 meters, which is promising for enriching data on OpenStreetMap.Camera Impact
The accuracy of Mapillary-generated data is influenced by the type of camera used. Cameras like the GoPro Hero 7 Black and GoPro Max performed better in terms of completeness and positional accuracy compared to the iPhone 11. This highlights the importance of GPS accuracy and image quality in data extraction.Recall Rate and Coverage
In a project evaluating traffic sign data, Mapillary generated map objects with more than a 95% average recall rate and an average horizontal positional accuracy of 2.16 meters. This indicates that Mapillary can effectively fill data gaps in OpenStreetMap, particularly for traffic signs.Limitations and Areas for Improvement
Despite its strengths, there are several limitations and areas where Mapillary can improve:Image Quality
Users have reported issues with image quality, such as dark, blurry, or tilted images, which can be useless for data collection. Implementing a quality score system for images could motivate users to capture better images.API Issues
The transition from API version 3 to 4 has introduced significant problems, including the loss of filters and information such as uploader and camera details. This has caused difficulties for developers and users, particularly those contributing to OpenStreetMap.Human Validation
While Mapillary’s AI algorithms are effective, there are cases where human validation is necessary to improve accuracy. The Mapillary Verifier Tool, which allows users to verify AI-detected objects, is a step in this direction but may need further development to address all accuracy issues.Engagement and Factual Accuracy
To enhance engagement and factual accuracy, Mapillary relies on community involvement. The Verifier Tool is a prime example, where users can help improve data accuracy by validating AI detections. This human-in-the-loop approach speeds up the learning process for the AI algorithms, leading to better data quality over time. In summary, Mapillary offers a valuable solution for geospatial data collection with good performance and accuracy, especially when using high-quality cameras and leveraging community validation. However, addressing issues with image quality and API functionality will be crucial for further improving its effectiveness. Mapillary - Pricing and Plans
Mapillary Pricing Structure and Plans
Mapillary’s pricing structure and plans are built around its unique model of providing crowdsourced street-level imagery and geographic data. Here’s a breakdown of the key aspects of their pricing and plans:
Revenue Streams and Plans
Mapillary generates revenue through several channels, but the primary focus is on licensing and subscription services.
Licensing Agreements
- Mapillary monetizes its enriched geographic data by offering it to enterprises and public sector organizations through licensing agreements. These clients, often in industries like logistics, transportation, and urban development, pay for access to the comprehensive mapping data.
API Subscription Service
- Developers can integrate Mapillary’s dataset into their applications through an API subscription service. This allows for the use of Mapillary’s extensive and up-to-date street-level imagery in various applications.
Free Options
- Mapillary provides free access to its map data globally through its APIs. Users can access map features and object detections via the Mapillary REST APIs without a subscription or cost. This includes the images API, the object detections API, and the map features API, all accessible with a single client ID.
Subscription Models
- While the core data is available for free, Mapillary offers various subscription models for more extensive use cases:
- SaaS Subscriptions: Part of their revenue model includes Software as a Service (SaaS) subscriptions, which likely cater to businesses and organizations needing regular access to updated mapping data.
- Enterprise Solutions: Customized solutions for enterprises that require more specialized and extensive use of Mapillary’s data.
- API Access Fees: Fees for accessing the API, particularly for larger volumes of data or more frequent usage.
Features Available
- Free Tier:
- Access to map features and object detections via APIs.
- Use of the Mapillary app for capturing and uploading images.
- Community-driven platform with real-time map updates.
- Paid Tiers:
- Access to enriched geographic data through licensing agreements.
- API subscription services for developers.
- Custom mapping services and enterprise solutions for more advanced needs.
Additional Costs
- While the primary data access is free, additional costs may apply for services such as custom mapping, higher API usage, or specific enterprise solutions. These costs are typically tied to the level of service and data usage required by the user or organization.
In summary, Mapillary offers a flexible pricing structure with free access to its core data through APIs, along with various paid subscription models and licensing agreements to cater to different user needs.
Mapillary - Integration and Compatibility
Mapillary Integration and Compatibility
Mapillary integrates seamlessly with a variety of tools and platforms, making it a versatile and widely compatible solution for managing and utilizing street-level imagery. Here are some key aspects of its integration and compatibility:
Integration with ArcGIS
Mapillary has a strong integration with ESRI’s ArcGIS tools, including ArcGIS Pro, ArcGIS Online, and Web AppBuilder. Users can access the Mapillary plugin from the ESRI marketplace and install it in ArcGIS Pro. This integration allows users to visualize Mapillary coverage, explore images, and extract vector data directly into their GIS projects. It also enables the overlay of global imagery on maps, providing seamless transitions from bird’s eye to street-level views.
Compatibility with Various Cameras
Mapillary supports imagery uploads from a wide range of devices, including smartphones, 360° cameras, action cameras, dashcams, and professional rigs. This flexibility makes it accessible to users with different types of equipment, whether they are using high-end cameras or everyday smartphones.
Mobile and Desktop Applications
Mapillary offers both mobile and desktop applications to facilitate the capture, upload, and management of street-level imagery. The mobile app is available for both iOS and Android, allowing users to capture and upload images on the go. The desktop uploader is available for macOS and Windows, enabling the upload of large amounts of images and videos.
Command Line Tools and API
For more advanced users, Mapillary provides command line tools for bulk image processing and uploading. Additionally, the Mapillary API offers vector tiles and entity endpoints, allowing developers to interact with Mapillary data in a flexible and scalable manner. This API supports various use cases, including rendering spatial data and accessing detailed properties.
OpenStreetMap and HERE Map Creator Integration
Mapillary imagery can also be integrated with OpenStreetMap (OSM) editing tools like iD editor and JOSM, helping users add detailed information to OSM. Furthermore, Mapillary has a built-in imagery layer in HERE Map Creator, providing a visual reference for editing HERE maps.
Privacy and Data Processing
Mapillary ensures user privacy through automatic blurring of faces and license plates detected in the images. The platform also uses computer vision and machine learning to process images, creating 3D views and extracting map data. This data can be accessed and shared in the cloud for free, enhancing the efficiency of mapping and GIS workflows.
Conclusion
In summary, Mapillary’s integration with various tools and platforms, along with its compatibility across different devices and cameras, makes it a comprehensive solution for collecting, processing, and utilizing street-level imagery. This integration enhances the capabilities of GIS tools like ArcGIS and supports community-driven data collection, making it a valuable resource for geospatial analysis and urban planning.
Mapillary - Customer Support and Resources
Customer Support Options
Mapillary provides several customer support options and additional resources to help users effectively utilize their platform.Reporting and Complaints
If you encounter any issues or need to report content that violates Mapillary’s Terms of Use, you can use the on-platform reporting tools. Alternatively, you can contact support directly via email at support@mapillary.zendesk.com. When reporting, it is helpful to provide a clear indication of the content’s location, a substantiated explanation of the issue, your name, and your email address.General Support
For any questions or concerns about the Mapillary Services or the Terms of Use, you can reach out to the support team at support@mapillary.zendesk.com. This is also the contact point for deleting your account or updating your account profile information.Feedback
Users can voluntarily submit feedback, questions, comments, or suggestions about Mapillary or its services. While this feedback can be used by Mapillary without compensation, it helps in improving the platform.Additional Resources
Mapillary offers various tools and integrations to enhance user experience:- Web, Mobile, and Desktop Apps: These tools allow users to capture, upload, and explore street-level imagery. The mobile app is available for both iOS and Android, and the desktop uploader supports uploading large amounts of images and videos.
- Command Line Tools: For bulk image processing and uploading, Mapillary provides command line tools that simplify the process.
- Integrations with Other Platforms: Mapillary integrates with popular mapping and GIS tools such as OpenStreetMap, ArcGIS, and HERE Map Creator, making it easier to use the imagery and map data within these systems.
Privacy and Data Management
For questions about how Mapillary handles personal data, users can refer to the Privacy Policy and contact the privacy team at privacy@mapillary.com. This includes information on data subject rights, such as accessing, correcting, or erasing personal data. By leveraging these support options and resources, users can effectively engage with the Mapillary platform and address any issues or questions they may have. Mapillary - Pros and Cons
Advantages of Mapillary
Mapillary offers several significant advantages that make it a valuable tool in the category of AI-driven search and mapping tools:Global Coverage and User Contribution
Mapillary allows anyone with a camera and an internet connection to contribute to a global database of street-level images. This crowd-sourced approach provides higher-density street view images, especially in areas where traditional mapping vehicles cannot reach.Frequency of Updates
The platform is constantly updated by its users, ensuring that the maps remain current and relevant. This frequent updating is particularly useful for applications such as public works, education, and autonomous driving.Detailed Street-Level Imagery
Mapillary’s images are captured from various perspectives, including walking, biking, and driving, offering a more comprehensive view of streets compared to traditional aerial or van-based mapping. This detail is enhanced by computer vision algorithms that help itemize objects and structures within each street scene.Automation and Computer Vision
The platform leverages advanced computer vision to automate the mapping process. Features include 3D reconstruction, semantic segmentation to label objects in images, and traffic sign recognition, which helps in creating accurate and detailed map data.Open Data and Collaboration
Mapillary promotes open data, allowing everyone to access and contribute to the mapping process. This collaborative approach helps in making better decisions about cities, roads, mobility, and navigation.Disadvantages of Mapillary
Despite its advantages, Mapillary also faces several challenges and limitations:Timely Updates and Privacy Concerns
One of the main drawbacks is the difficulty in ensuring timely updates across the entire globe. Additionally, while algorithms and manual reviews are used to protect privacy by blurring faces and license plates, there can still be delays and instances where sensitive information is not immediately blurred.Image Quality Issues
Capturing images, especially while biking or in less ideal conditions, can result in blurred or out-of-focus photos. The app’s settings and the quality of the camera mount can also affect image quality, leading to a significant number of unusable images.Map Loading Times and Interface
The high density of images in some areas can lead to slow map loading times. The interface also lacks a satellite view, which might be unfamiliar to users accustomed to Google Earth or similar services.Technical and Support Issues
Users have reported issues with the app’s performance, such as problems with image rotation, compass direction recording, and the lack of a structured issue tracker for reporting bugs and feature requests. This can make it harder for users to track the status of their reported issues. Overall, Mapillary offers a unique and valuable approach to mapping by leveraging crowd-sourced contributions and advanced computer vision, but it also faces challenges related to image quality, privacy, and user experience. Mapillary - Comparison with Competitors
Mapillary’s Unique Features
- Mapillary is a street-level imagery platform that generates map data through collaboration and computer vision. It allows users to upload pictures from various devices, leveraging consumer-grade cameras to create high-quality map data. This approach eliminates the need for expensive survey vehicles, making it more accessible and cost-effective.
- Mapillary’s technology is device-agnostic, enabling data collection from a wide range of sensors and compensating for any shortcomings of consumer-grade devices. This flexibility is enhanced by the company’s investment in developing advanced computer vision algorithms.
Competitors and Alternatives
HERE Technologies
- HERE Technologies specializes in location technology and spatial intelligence, offering a platform for building, deploying, and scaling location solutions. Unlike Mapillary, HERE Technologies focuses more on enterprise-level location services and does not rely on crowdsourced imagery.
Propeller
- Propeller provides drone technology and data solutions primarily for the construction, mining, and waste management sectors. While Propeller also uses aerial imagery, its focus is more on specific industrial applications rather than general street-level mapping.
Civil Maps
- Civil Maps offers 3D mapping technology for fully autonomous vehicles, using artificial intelligence and vehicle-based local processing. This is more specialized towards the automotive industry, particularly for autonomous driving, unlike Mapillary’s broader application.
DeepMap and DeepMotion
- DeepMap and DeepMotion focus on mapping and localization solutions for autonomous vehicles. DeepMap offers high-precision mapping, while DeepMotion provides crowdsourced HD maps with localization and mapping at 10cm accuracy. These solutions are more tailored to the automotive sector and autonomous driving, differing from Mapillary’s general street-level imagery approach.
CARMERA
- CARMERA develops a street-level intelligence platform providing 3D maps and navigation data for autonomous vehicles. While it shares some similarities with Mapillary in terms of street-level data, CARMERA’s focus is more on autonomous vehicle applications.
Xverum and SafeGraph
- Xverum and SafeGraph are data companies that provide structured datasets and high-precision location data, respectively. They do not focus on street-level imagery but rather on providing clean and structured data for various applications, making them different from Mapillary’s visual data collection.
OpenStreetMap
- OpenStreetMap is an open-source mapping project that relies on community contributions to create and maintain map data. While it shares Mapillary’s collaborative approach, OpenStreetMap does not use AI-driven computer vision to the same extent and is more focused on community-driven mapping.
Mapbox
- Mapbox offers APIs, SDKs, and live updating map data for developers to create mapping, navigation, and search experiences. Mapbox Studio provides extensive customization options but does not focus on street-level imagery collection like Mapillary. Instead, it leverages existing map data to enhance user experiences.
Key Differences and Considerations
- Industry Focus: Mapillary’s neutrality and openness allow it to partner with various mapping and navigation companies without competition, making it a versatile option. In contrast, many competitors have more specialized focuses, such as autonomous vehicles or industrial applications.
- Data Collection: Mapillary’s use of consumer-grade devices and AI-driven computer vision sets it apart from competitors that may require more specialized equipment or focus on different types of data.
- Community and Collaboration: While OpenStreetMap is also collaborative, Mapillary’s integration with computer vision algorithms provides a unique blend of community-driven data collection and advanced technological processing.
When choosing an alternative to Mapillary, consider the specific needs of your project:
- For autonomous vehicle applications, Civil Maps, DeepMap, or CARMERA might be more suitable.
- For industrial or construction-focused mapping, Propeller could be a better fit.
- For general mapping and navigation with extensive customization, Mapbox might be the way to go.
- For open-source and community-driven mapping, OpenStreetMap is a strong option.
Each of these alternatives offers unique features and focuses that can cater to different requirements and use cases.
Mapillary - Frequently Asked Questions
Frequently Asked Questions about Mapillary
Can I upload single photos to Mapillary?
While you can technically upload single photos, it is recommended to capture sequences of photos. This is because having photos taken close to each other is crucial for creating 3D reconstructions of places and providing a better visual experience when exploring the photos.
Can I take photos in portrait orientation?
No, Mapillary supports landscape orientation only. It is important to keep your camera or phone horizontally when capturing photos to ensure they are compatible with the platform.
How long does it take for my photos to appear on the map?
After uploading your photos, they will be processed by Mapillary’s computer vision technology to blur faces and license plates and connect them with other photos in the same area. This process typically takes a few hours, but you should see your uploads within a day. If you upload videos, the processing time will be longer.
Why are my tracks on the map zig-zagging?
The tracks on the map are created by recording location information alongside your photos, usually using a device with GPS. The precision of GPS can vary, especially with everyday devices, leading to tracks that may wobble around the actual path taken. You can correct these tracks post-upload.
How does Mapillary handle privacy concerns, such as faces and license plates in photos?
Mapillary uses computer vision technology to detect and blur faces and license plates on the photos before they are visible on the platform. The detection algorithms are continuously improved, and there is also a system for manual correction of blurs if needed.
Does Mapillary own the photos I contribute?
No, you retain the full rights to the photos you contribute. Mapillary is granted the rights to use the photos, but ownership remains with you.
What cameras are recommended for capturing 360° photos on Mapillary?
For 360° capturing, Mapillary recommends using cameras like the LG 360, Ricoh Theta S, and Giroptic 360. These cameras capture decent-quality images with high frequency and have integrations with the Mapillary app to make capturing easier.
Can I contribute photos using any camera?
Yes, you can contribute with any camera, as long as you can record the location information that needs to accompany the images. This can be done using a smartphone with the Mapillary app, an external camera with GPS, or by combining photos with GPS tracks from another device.
How does Mapillary’s Time Travel feature work?
The Time Travel feature works based on the location and direction information captured by your device’s GPS and compass. If the data does not match precisely enough, the Time Travel function may not activate even if you are in the same spot.
What are the main uses of Mapillary images beyond just viewing places?
One of the main uses of Mapillary imagery is to edit geospatial data. The platform allows users to extract valuable information from ground-level views, either visually or through automated computer vision. This is particularly useful for updating GIS data.
How does Mapillary enhance map quality using computer vision?
Mapillary enhances map quality by using computer vision to detect and map various features such as utility poles, street lights, mailboxes, and traffic signs. This technology helps in creating hyper-detailed maps, which are essential for applications like autonomous vehicles.
Mapillary - Conclusion and Recommendation
Final Assessment of Mapillary
Mapillary is an innovative platform that leverages crowd-sourced street-level imagery to create and update detailed maps. Here’s a comprehensive assessment of its value and who would benefit most from using it.Key Features and Benefits
Crowdsourced Imagery
Mapillary allows users to upload street-level photos taken with any camera, which are then integrated into a cohesive map database. This approach ensures continuous updates and a wide coverage of geographic areas.
Geotagging and Map Creation
Images are automatically geotagged, and the platform uses computer vision technology to synchronize and merge photos from different times and contributors, creating detailed and up-to-date street-level maps.
Community Interaction
Users can interact through comments and likes on images, fostering a sense of community among contributors.
Data Analysis and API Access
Mapillary offers tools for analyzing street-level imagery and extracting data such as traffic signs or building facades. Developers can access the platform’s API to integrate this data into their applications.
Privacy Measures
The platform employs advanced blurring techniques to protect personal identities and sensitive information in the photos, ensuring user privacy.
Who Would Benefit Most
Mapillary is highly beneficial for several groups:Urban Planners and Government Bodies
These entities can use the real-time, high-resolution imagery for urban development initiatives, such as planning transportation layouts, managing public utilities, and designing pedestrian-friendly spaces.
Geospatial Professionals
Individuals working in GIS applications can utilize Mapillary’s detailed and up-to-date street-level imagery to enhance their mapping projects.
App Developers
Developers can integrate Mapillary’s comprehensive dataset into their applications through the API, enhancing the accuracy and functionality of their services.
Transportation Agencies and Logistics Companies
These organizations can benefit from the up-to-date street-level imagery for route planning, infrastructure management, and other related tasks.
Environmental Researchers and Real Estate Firms
These groups can use the detailed imagery to analyze environmental changes or assess property conditions.