Forestry

Forest environments are critically important to the climate change equation, a major source of global economic activity, and home to innumerous species of flora and fauna.

Advances in technology have made it possible to better understand and utilize forest resources in a sustainable manner. In particular, LiDAR scanning technologies and high-resolution satellite and airborne imagery have radically changed the science of forest management.

eCognition software provides foresters with the ability to conduct rapid, precise, large-scale forest analysis. Data sources are fused and analyzed to automate forestland classification, conduct tree species identification, quantify forest canopies and biomass, delineate individual tree crowns, and measure and monitor firescars or cutblocks. 

Typical Applications

  • Quantify canopies and biomass
  • Identify species
  • Delineate tree crowns
  • Measure fire scars and cutblocks

Urban

Urban environments are home to approximately 50% of the world’s population – a percentage that increases every year.

The existence of more, and larger, cities creates an enormous challenge for urban planning, governance, and renewal.  Airborne surveys, high resolution satellite imagery, and LiDAR scanning provide authorities improved opportunities to effectively plan and govern the modern urban environment.

eCognition software enables state and local governments, city planners, and transportation and environmental authorities to import, fuse, interpret, and analyze data from different sources to achieve a greater understanding of their urban environments. The results include improved accuracy in taxation, effective transportation and infrastructure design, and healthier urban environments more in balance with nature.

 

Typical Applications

  • Plan infrastructure
  • Assess property
  • Improve taxation
  • Build greener cities

Diverse Landcover

Landcover assessments are one of the most common application areas for eCognition software. Our object-based approach uses context information to automate classification and quantification of diverse environments - an approach which has been repeatedly proven to be highly effective and efficient for landcover projects in a wide variety of environments.

The ability to develop production workflows by combining products from the eCognition software suite means that landcover projects of vitually any scale can be completed with the highest levels of accuracy, from individual studies to nation-wide anaylsis.

 

But does context really matter?



Question: If the maple trees in Central Park are spectrally identical to those in the forests of New York State, are they the same?

Answer: Not if you only want to meaure urban parkland. Intelligently automating the classification of diverse landcover environments requires far more than a simple classification based on sample pixels – it requires context. Learn more

Typical Applications

  • Transport corridor analysis
  • Pipeline or communication infrastructure planning
  • Mapping ecosystems
  • Enviromental change detection
  • Uban growth evaluation and planning

Agriculture

The combination of advanced aerial and satellite-based sensors with GPS and GIS has transformed agriculture monitoring and precision farming into a major remote sensing application area.

Very high resolution aerial and satellite imagery contributes to effective time management and resource usage at the farm-level. Equally as important is the understanding of interactions between agriculture and surrounding environments to ensure a sustainable balance.

eCognition software utilizes context to intelligently classify and quantify image and vector data for analysis of individual farms and vast farming regions – a significant advantage for those using imagery for large-scale agricultural monitoring. Applications include verification of farming subsidy applications, assessment of insurance claims, and crisis response in cases such as disease,  parasite outbreak, or flood conditions.

 

Typical Applications

  • Evaluate land use and subsidy claims
  • Regional resource management
  • Assess crisis response
  • Incorporate images into percision farming

Marine and Riparian Zones

Coastal and riparian environments are highly vulnerable to natural and human-driven changes. Tools and procedures to study and monitor ecosystems, manage national boarders, and rapidly respond to disasters such as hurricanes, tsunami, and oil spills, are essential. Satellite multispectral digital image data are especially helpful for monitoring and detecting change, improving harbour management, and directing security and disaster response.

eCognition's advanced context-based architecture enables an improved understanding and mapping of important ecosystem parameters, identification of land and water-based forms and their changes, as well as urban growth patterns in coastal regions. The results enable better decisions about land and resource use, management practices, and coastal zone and watershed development.

 

Typical Applications

  • Ecosystem study and monitoring
  • Harbour and boarder managment
  • Rapid disaster response

Atmosphere and climate

Data for climate analysis can be obtained through microwave or radar sensing, which is ideal for observation regardless of light or weather conditions. This provides an uninterrupted flow of accurate data for interpretation, and long-term observations can be compiled to monitor climate changes. Therefore, the combination of data types, usage and analysis techniques in this area are varied.

Of course, atmosphere is also an important and costly hinderance for collection of satellite data, specifically when clouds obsecure important image data. Locating the cloud areas within image data is an important and valuable task which can be automated using Definiens software.

Typical Applications

  • Cloud detection
  • Climate / environmental change studies
  • Analysis of radar data

Defense and Security

eCognition software offers some distinct advantages for defense and security applications. Rapid mapping and data fusion  provide detailed maps of any territory, however remote or unfamiliar. Context-based analysis links object detection with feature extraction. Targets can be detected and put into relation to understand the depicted situation.

Of critical importance is  eCognition's capability to utilize, integrate with, or interface to many different sensors and systems. For example, satellite and ground radar systems, vessel monitoring systems using GPS, and many types of airborne remote sensors may provide data that needs to be fused, analyzed, and exported to another system. These capabilities help eCognition deliver world class geospatial analysis capability for defense and security, enabling timely and well-informed decision making.

Typical Applications

  • Rapid mapping of unfamiliar territory
  • Identification and tracking of objects of interest
  • Unmanned vehicles navigation (atypical, but demonstrated)
  • Change detection on sensitive facilities
  • Generation of 3D simulation databases
  • UAV image processing
  • Data fusion

Airborne Optical

Aerial photography has undergone some fundamental changes in recent years. Today very high resolution digital images of outstanding quality and precision can be produced and delivered within very short time frames.

eCognition is specifically designed to analyze high resolution data, offering the unique capability of fusing data from various sources. Therefore aerial photography can be combined with digital elevation models derived from stereo imagery or LiDAR, satellite imagery as well as cadastral vector layers. Leveraging this capability, Definiens customers optimize their available data to bring innovative solutions to market.

 

Efficient Workflows

  • By fully integrating into existing workflows, both processing and editing can be streamlined, resulting in higher automation and throughput.


Leveraging Data Synergies

  • By applying an object-based approach, eCognition overcomes the boundaries of differing data types, sources and resolutions. Data from optical sensors, elevation data, and GIS vector information can be fused; leading to more detailed and accurate of results.


Building Solutions

  • eCognition software provides all components necessary to build solutions tailored to the specific needs of a subject matter expert. This allows additional revenue to be generated based on airborne imaging data.

 

Typical Applications

  • High resolution landcover mapping
  • Building detection
  • Impervious surface mapping
  • Change detection
  • Forest Inventory

Hyperspectral

Hyperspectral imaging collects and processes information from across the electromagnetic spectrum. The sensors used look at objects using a vast portion of the electromagnetic spectrum resulting of datasets containing a vast amount of image layers representing the different spectral bands.

eCognition software offers unique advantages making it a perfect match for hypespectral data. The unique capability of feasibly using data from various sources allows combining hyperspectral data with additional data such as LiDAR, Radar or GIS data. Further, the rule based classification approach allows the wealth of information contained in Hyperspectral imagery to be leveraged in an optimal way. With this capability, Definiens' customers optimize their available data to bring innovative solutions to market.

 

Typical Applications

  • Agricultural land use mapping
  • Forest inventory
  • Vegetation mapping

LiDAR

With the emergence of LiDAR as an exciting new data format, eCognition has proven to be perfectly suited to the extraction of intelligence from LiDAR data by a global user community.

eCognition offers the unique capability of fusing data from various sources. Therefore LiDAR can be combined with other available data like aerial photography, satellite imagery as well as cadastral vector layers. Leveraging this capability,  eCognition customers optimize their available LiDAR data to bring innovative solutions to market.

  • Efficient Workflows
    By fully integrating into existing workflows both processing and editing can be streamlined, resulting in higher automation and throughput.

  • Leveraging Data Synergies
    By applying an object-based approach, eCognition overcomes the boundaries of differing data types, sources and resolutions. Data from optical sensors, LiDAR and GIS vector information can be fused leading to more detailed and higher accuracy of results.

    Building LiDAR Solutions
    eCognition software provides all components necessary to build solutions tailored to the specific needs of a subject matter expert. This allows additional revenue to be generated based on LiDAR data.

 

Read about using LiDAR with eCognition in GeoConnection: 
"Object Based Image Analysis and LiDAR"

 

 

Case Studies

    More

Typical Applications

  • Impervious surface mapping
  • Forest Inventory
  • Building detection

Multispectral

Most remote sensing data taken by satellite can capture light from frequencies beyond the visible light range, such as infrared. This facilitates the extraction of additional information that the human eye fails to capture with its receptors for red, green and blue.

eCognition software offers a number of advantages making it a perfect match for multispectral data. By following an approach in which data can be segmented on individual image layers without restricting the following classification steps, users can leverage the specific advantages of the different spectral bands within a multispectral dataset.

The rule based classification approach allows the wealth of infromation contained in multispectral imagery to be leveraged in an optimal way. With this capability, eCognition customers optimize their available data to bring innovative solutions to market.

 

 

Typical Applications

  • Agricultural land use mapping
  • Forest inventory
  • Vegetation mapping

Radar

While radar technology does not match optical remote sensing data in resolution and detail sharpness, it has other advantages. Depending on the frequency, the radiation penetrates cloud cover or even the ground surface including vegetation and soils.

Due to its low signal to noise ratio, radar data is very hard to analyze. Thanks to its object oriented analysis approach, eCognition is a perfect match for radar data. By performing analysis on objects rather than pixels, the texture contained in radar data as well as the shape of areas can be leveraged for classification.

In addition, eCognition offers the unique capability of fusing data from various sources. Therefore radar can be combined with other available data such as aerial photography, satellite imagery as well as cadastral vector layers. Leveraging this capability, eCognition customers optimize their available radar data to bring innovative solutions to market.

 

  • Efficient Workflows

    By fully integrating into existing workflows both processing and editing can be streamlined, resulting in higher automation and throughput.

  • Leveraging Data Synergies

    By applying an object-based approach, eCognition overcomes the boundaries of differing data types, sources and resolutions. Data from optical sensors, LiDAR and GIS vector information can be fused leading to more detailed and higher accuracy of results.

  • Building Solutions

    eCognition  software provides all components necessary to build solutions tailored to the specific needs of a subject matter expert. This allows additional revenue to be generated based on radar data.

Technology Behind

The underlying technological principle eCognition is that it is context-based. The technology identifies objects rather than just examining individual pixels. It then makes inferences about those objects by looking at them in context.

 

eCognition makes a radical departure from conventional approaches to image analysis due to it's ability to emulate the human mind's cognitive powers. Using patented segmentation and classification processes, we have developed a robust method of rendering knowledge in a semantic network. The technology examines pixels not in isolation, but in context. It builds up a picture iteratively, recognizing groups of pixels as objects. Just like the human mind, it uses the color, shape, texture and size of objects as well as their context and relationships to draw the same conclusions and

inferences that an experienced analyst would draw. Though somewhat simplified, the following example illustrates the basic principles.

With the addition of a knife and fork, the circle is immediately classified as a plate. The object the circle hasn't changed, but the human mind uses context and relationships with other objects to make intelligent inferences about the objects it perceives. And this is precisely how eCognition works.

eCognition Networks

eCognition classifies and analyzes imagery using all the semantic information required to interpret it correctly. Rather than examining individual pixels in isolation, it distills meaning from the objects' connotations and mutual relations.

Of course, it also recognizes pixel information. However, it adds significant value to this information by creating a powerful cognition network in a series of iterative segmentation and classification steps. This dramatically enhances the value of intelligence and information extracted from the image. What's more, the extracted information is fully quantified and qualified to meet the user's specific demands.

Customers