Table of Contents
Why Image Blurring Solution for Post-Processing
In this digital age and Internet of Things, private data protection, synonymously known as data privacy, holds utmost importance. As the facial recognition technologies become smarter, the need to enforce meticulous blurring technologies becomes urgent. Face and Body form the most important elements of Personal Identification Information (PII). High resolution mobile mapping systems capture the imagery such as faces, body, license plates which conflict with General Data Protection Regulations (GDPR) and other data privacy laws.
Private data in NavVis Ivion
Advanced mapping solutions capture a wider range of reality and the existing site-conditions, objects, and people in incredible resolution. When these dense point clouds are navigated, they almost look like a video with clear PIIs. This presents serious privacy concerns in accordance with regulations.
In surveying, the ideal scanning and mapping conditions almost never exist. On construction sites and functioning facilities, it is difficult to evict the workforce, as the disruption costs a lot of time and effort to dismiss and then retrieve the operations.
Anonymization is a critical step, as further web processing depends on the speed of image processing and directly impacts the cost and processing time. Using an on-premises (offline) anonymization software solution ensures privacy, data-security, and accessibility. Thus, it is imperative to carry on the surveying with the on-site conditions. The PII captured of the people is processed through this state-of-the-art blurring solution, that makes the mobile mapping solutions compliant with GDPR & other data protection regulations.
The use case for the automated detection feature of this tool could be extended to locate building materials, furniture, and electronic devices, for example, in a facility or in a construction site. This could evolve to automatically assign points of information to objects and link them in a data group. Numerous use cases could be developed on this automated detection feature of objects, for counting and keeping track of objects with points of information (PoIs).
Anonymization Solution for NavVis Mobile Mapping Systems
Features of Pano-Blurring Solution:
- Automated anonymization using machine learning
- 99% success rate for detection and blurring
- Applications in both panorama and planar images
- AI powered
- GDPR compliant
- Highly scalable
- No false detections
DiConneX (in collaboration with Trexcare) has developed a Pano-Blurring Tool that works in synchronization with the data post-processing software such as NavVis SiteMaker . The raw data from advanced surveying and mobile mapping consists of 360 ° images, which is a combination of panoramic and planar images in .jpg or .png extension. These form the input for the first stage of Post Processing software such as NavVis Sitemaker. The open-source AI powered algorithm, EfficientDet, automatically anonymizes the predefined personal identification elements. The software can achieve anonymization at an impressive speed, even in difficult lighting conditions.
EfficientDet algorithm is based on COCO (Common Objects in Context) dataset. These datasets enable higher precision for the detection and recognition of defined objects in a model. The technology uses COCO datasets for facial, body, and object detection in images, and in models, which will evolve from image blurring to selective blurring of personal objects, such as license plates, ID cards, etc. Blurring is the default style of image processing for anonymization in large image datasets, which is used by companies such as Google in Google Street View. These objects and persons are identified not only in isolation, but also in their natural habitat or environment, eg, on a busy street, a train station, or a shopping mall. This AI technology blurs out sensitive information such as logos, photographs, vehicles, and so on. With the Blurring Tool,
This tool was tested in the mapping project for Hamburg Hauptbahnhof, which is one of the busiest locations in Hamburg with a footfall of 537,000 per day. As a result, the tool delivered a 96% -99% success rate, and the inaccuracy was due to any obstruction on the face or partial capture of the body in the mapping.
Processed Image from Hamburg-Hauptbahnhof (Link to HH-HBF Digital Twin Youtube Video)
The data-driven analyzes of three areas – people, tooling, and process – have shown promise in improving productivity compared to manual blurring, which has lower accuracy and with further applications across industries. It is a cost-effective way to comply with data protection regulations.
Privacy concerns are a major threat to the growth of mobile mapping systems. The hurdle lurks in the form of privacy breach, in an environment of GDPR raise of data-privacy awareness. Surveying companies, by law, need to process pictures before proceeding to spatial data platforms such as NavVis Ivion. The processing should be scalable for large image datasets, secure and accessible anytime, where you need it.
DiConneX offers a fully automated, efficient, and reliable blurring solution to anonymize large image datasets with GDPR-compliant data-processing systems. It supports various image formats and uncomplicated handling, which is ideal for NavVis Core and NavVis Enterprise Partners. For a free demo of the Image Blurring solution, Use Cases across industries, or any further questions, please contact us using the form below: