吃晚饭porting artificial intelligence solution to parking problem

Author:琳达·达菲（Linda Duffy）

With a finite number of parking spaces and an increasing number of cars, frustration and inconvenience related to parking is growing. Recognising this as a problem, Manuela Rasthofer, CEO ofTerraLoupe GmbH, launched a project to结合人工智能和矫形空中图像在整个德国创建可用的停车场和空间的准确清单。

Mapping available parking

停车汽车可能是一项压力且耗时的活动，将来，自动导航的车辆将在没有驾驶员提供帮助的情况下寻找空间。需要准确测量和识别包括停车位的所有类型对象的高清数字地图的需求迅速成为现实。

TerraLoupe GmbH is a technology start-up based in Munich, Germany, that focuses on combining geodata and computer analytics in innovative ways. Starting with high-resolution orthoimagery, TerraLoupe appliesmachine learning algorithms to detect and measure objects in the physical world，例如建筑物，道路和树木，以创建数据丰富的3D模型。

“To address the growing parking problem, we wanted to see if it was feasible to detect and assess parking lots using aerial imagery and artificial intelligence,”Rasthofer说。“By automating the extraction of features and digital content, we thought we could greatly reduce the time it took to create maps, without sacrificing the accuracy.”

Acost-effective method of creating digital maps对于一级汽车供应商和原始设备制造商（OEM）来支持自主导航行业，这特别有趣。但是，许多其他行业也可以利用这些信息。

HXGN内容程序提供

Semantic lane model created out of detected lane markings based on 15 centimetres GSD data in Martinsburg

In 2014, theHXGN内容程序开始收集投机现成的orthorectified imageryof the US, parts of Europe, and populated areas of Canada tocreate a database available to customers。这goal was to acquirecloud-free 30-cm resolution, 4-band imageryover less populated areas, and15厘米分辨率over metro areas with a人口大于50,000。

通过HXGN内容程序，Terraloupe获得了15厘米分辨率orthoimagery of Berlinto test its internally developed object-identification algorithms. The initial work on Berlin tookeight weeks to train the algorithms准确识别和分类停车位，其次是three days to analyse and produce mapsfor all of Germany.

“Access to imagery through the HxGN Content Program允许我们download the geographic locationswe need, and thentrain our algorithms on the new data，”Rasthofer解释了。“这re are always slight differences in architecture, infrastructure, and road systems that are unique to each country. We check the confidence interval for each object and recheck low percentages. As we correct errors, the algorithms continue to learn and improve until we achieve a very high accuracy level.”

这可通过HXGN内容程序获得航空矫形图进行严格的QA/QC过程，以确保提供调查级图像。“这HXGN内容程序best suits the needs of our customers in the自动驾驶，停车援助和损失报告的领域for insurance/reinsurance companies,”Rasthofer说。“我们也成功交付intelligence related to infrastructure, utilities, railways, and others for a variety of purposes.”

机器学习加快了准确的映射

Terraloupe的项目表明high-resolution aerial orthoimages combined with machine learning可以有效地用于提取数字内容。停车分析提供了有用的信息，例如停车场的位置，入口和出口，以及可以适合每个批次的不同类别（紧凑，中型，大型）的汽车数量。城市规划师，送货人员，出租车司机和拥挤零售区的顾客都可以从这种改善的停车情报中受益。

“这availability ofhigh-resolution, high-accuracy imagerydetermined where we started the project; however, we intend to perform this analysis on all of Europe as data becomes available through Hexagon, and we’d like to expand our services into the U.S.,”Rasthofer说。“Overall our goal is to efficiently extract all types of objects and创建一个完整的数字环境。”

获得aerial imagery is faster和更多efficientthan terrestrial methods, allowing more frequent updates, which is crucial for many applications. Hexagon’s global operations generate widespread availability of imagery and good business partnerships with data providers to continue to meet thegrowing demand for digital maps。