Streamline Your Workflow: FME & Picterra Integration

Table of Contents

1. Introduction
2. Understanding FME and Pikterra Integration
3. The Importance of Computer Vision in Facial Data Detection
4. Application in Aerial Imagery Analysis
5. Analysis of Runway Conditions at Vancouver International Airport
   1. Identifying Areas of Interest
   2. Uploading Images to Pikterra
   3. Training the Crab Detector
   4. Defining Training Areas and Tracing Cracks on the Pavement
   5. Running the Training and Assessing the Results
6. Detection on a Larger Scale with FME and Pikterra
   1. Creating a New Workspace
   2. Requesting Roster and Detector Information from Pikterra
   3. Preparing the Data for Output
   4. Running the Detection and Receiving Detected Objects
7. Generating Reports and Work Orders
   1. Reading the Detected Cracks Data Set
   2. Calculating the Percentage of Area Covered with Cracks
   3. Overlaying Cracks on Images for Reports
   4. Generating Excel Spreadsheets and Multi-page PDF Reports
   5. Submitting Work Orders for Areas with Excessive Cracks
8. Conclusion

Understanding FME and Pikterra Integration

Computer vision technology has revolutionized the way we analyze visual data. In this article, we will explore the integration between FME and Pikterra, a computer vision web platform designed specifically for detecting objects on imagery taken by drones, planes, or satellites. We will Delve into the application of this integration in the Context of analyzing runway conditions at Vancouver International Airport. Through a step-by-step guide, we will showcase how FME and Pikterra work together to identify areas of interest, upload images, train detectors, run detections, generate reports, and prepare work orders for fixing problems in the pavement. Whether You are a GIS professional or simply interested in the intersection of computer vision and geospatial analysis, this article will provide valuable insights into the capabilities of FME and Pikterra.

The Importance of Computer Vision in Facial Data Detection

Computer vision has become an integral part of numerous industries, including facial data detection. With the ability to accurately identify and analyze objects on imagery, computer vision technology opens countless opportunities for improving efficiency and decision-making. In the case of facial data detection, it enables applications such as facial recognition, emotion detection, and object tracking. With FME and Pikterra integration, businesses can harness the power of computer vision to enhance their geospatial analysis capabilities and gain valuable insights from aerial imagery.

Application in Aerial Imagery Analysis

Aerial imagery analysis plays a crucial role in various fields, from urban planning and infrastructure management to environmental monitoring and disaster response. By leveraging the integration between FME and Pikterra, organizations can streamline their analysis workflows and extract actionable information from aerial imagery more effectively than ever before. In the following sections, we will explore a specific use case of this integration - analyzing runway conditions at Vancouver International Airport.

Analysis of Runway Conditions at Vancouver International Airport

1. Identifying Areas of Interest

Before diving into the analysis, it is crucial to identify the areas of interest that require assessment. In the case of Vancouver International Airport, We Are particularly concerned about the condition of the runways and taxiways. By using computer vision technology, we can detect cracks in the pavement and focus our analysis on areas where potential issues might arise.

2. Uploading Images to Pikterra

To initiate the analysis, we need to upload the Relevant imagery to Pikterra. The integration with FME allows us to seamlessly transfer image data to Pikterra's cloud storage. By providing the necessary details, such as the transport mode, API key, and roster names, we can ensure that our images are securely uploaded for further processing.

3. Training the Crab Detector

Training the crab detector is a crucial step in improving the accuracy and reliability of crack detection. By adding more samples to the existing detector and incorporating image 8 along with two other images, we can enhance the detection capabilities. With Pikterra's user-friendly interface, training the detector becomes a seamless process, empowering users to refine their models and achieve more accurate results.

4. Defining Training Areas and Tracing Cracks on the Pavement

In order to train the detector effectively, we need to define training areas and Trace cracks on the pavement. This involves specifying areas without cracks, as well as assessment areas that will be used to evaluate the detector's performance. By leveraging the tools provided by Pikterra and using GIS software like FME, we can precisely mark these areas and Gather the necessary training data.

5. Running the Training and Assessing the Results

After defining the training areas and tracing cracks, we can proceed with running the training. This process involves feeding the data into Pikterra's training pipeline and allowing the system to analyze and learn from the provided information. Once the training is complete, we can assess the results by inspecting the preview in the assessment areas. This step helps us gauge the effectiveness of the training and determine if additional iterations are required.

Detection on a Larger Scale with FME and Pikterra

1. Creating a New Workspace

Having successfully trained the crab detector on a smaller scale, we can now shift our focus to detecting cracks on a larger scale. This requires the creation of a new workspace in FME, where we can orchestrate the transformation and integration with Pikterra.

2. Requesting Roster and Detector Information from Pikterra

To kickstart the detection process, we need to request the list of all the rosters uploaded to Pikterra's service. Additionally, we retrieve information about the detectors available. In this case, we specifically require the detector for cracks. By obtaining this essential information, we can proceed with the detection process seamlessly.

3. Preparing the Data for Output

Before running the detection, it is essential to prepare the output data for our destination system. This involves setting up the file geodatabase Writer, defining the feature Type, and specifying the attributes we want to include in the output. By configuring these settings in FME, we ensure that the data is compatible with our desired format and can be easily utilized for further analysis.

4. Running the Detection and Receiving Detected Objects

With all the necessary preparations in place, it is time to run the detection process. By utilizing Pikterra's robust detection capabilities and leveraging the integration with FME, we can efficiently detect objects of interest in the provided imagery. Once the detection is complete, we receive the detected objects, which we can further process and analyze.

Generating Reports and Work Orders

1. Reading the Detected Cracks Data Set

To generate reports and work orders Based on the detected cracks, we begin by reading the complete data set. This includes all the cracks that have been detected in the pavement. By accessing this data, we can proceed with the necessary calculations and analysis.

2. Calculating the Percentage of Area Covered with Cracks

One important aspect of assessing cracks is determining their impact on the overall pavement. By calculating the percentage of the area covered with cracks, we can gauge the severity of the issue. This calculation helps prioritize areas that require immediate Attention and further analysis.

3. Overlaying Cracks on Images for Reports

To Visualize the extent of cracks, we overlay the detected cracks on top of the original images. This provides a comprehensive view of the cracks' location and distribution, aiding in the generation of accurate reports and work orders. By leveraging FME's geospatial capabilities, we can effectively overlay the detected cracks and Create visual representations for analysis.

4. Generating Excel Spreadsheets and Multi-page PDF Reports

To facilitate data analysis and presentation, we generate Excel spreadsheets and multi-page PDF reports. These reports contain essential information about each area of interest, including the total area covered with cracks and the percentage of coverage. We enhance the reports by using color coding to indicate the severity of the cracks. This comprehensive documentation helps stakeholders understand the situation and make informed decisions.

5. Submitting Work Orders for Areas with Excessive Cracks

In cases where the area covered by cracks exceeds a predefined threshold, we generate work orders for necessary repairs. This process involves integrating with systems like City Works to efficiently submit work orders and initiate the necessary actions. By automating this workflow, we ensure that areas with excessive cracks are promptly addressed, minimizing risks and ensuring the safety of the runways and taxiways.

Conclusion

In this article, we explored the integration between FME and Pikterra and its application in analyzing runway conditions at Vancouver International Airport. By leveraging computer vision technology, organizations can enhance their geospatial analysis capabilities and improve decision-making processes. The step-by-step guide provided insights into the workflow, from identifying areas of interest to generating reports and work orders. With FME and Pikterra integration, users can efficiently analyze aerial imagery, detect objects of interest, and take proactive measures to address potential risks. Whether you are involved in infrastructure management, urban planning, or environmental monitoring, this integration offers a powerful toolset to enhance your analysis workflows and make data-driven decisions.

Highlights

• Integration between FME and Pikterra empowers geospatial professionals to leverage computer vision technology for analyzing aerial imagery.
• By training detectors and running detections, organizations can accurately detect cracks in pavement and assess runway conditions.
• Generated reports and work orders help prioritize repairs and maintenance, enhancing overall runway safety.
• FME's seamless data integration capabilities and Pikterra's user-friendly interface make the workflow efficient and user-friendly.
• The integration between FME and Pikterra unlocks the potential of computer vision for enhanced geospatial analysis and decision-making processes.

FAQ

Q: Can FME and Pikterra integration be used for tasks other than crack detection on runways? A: Absolutely! FME and Pikterra integration can be utilized for various tasks, including object detection in aerial imagery for infrastructure inspection, environmental monitoring, and land use analysis.

Q: Is computer vision technology limited to aerial imagery analysis? A: Not at all! Computer vision technology can be applied to different types of imagery, including satellite imagery, drone imagery, and even street-level imagery. Its applications range from agriculture and disaster response to retail and security.

Q: How does FME facilitate the integration with Pikterra? A: FME acts as the orchestrator in the integration process, seamlessly transferring data between systems, preparing it for analysis, and generating outputs in the desired format. FME's data transformation capabilities make it an ideal tool for integrating with Pikterra's computer vision platform.

Q: Can I use the generated reports and work orders in my existing GIS software? A: Absolutely! The output from the FME and Pikterra integration can be easily imported into your preferred GIS software, allowing you to incorporate the generated reports and work orders into your existing workflows.

Q: Does the integration require extensive programming knowledge? A: No, FME and Pikterra integration can be achieved without extensive programming knowledge. Both systems offer intuitive interfaces and workflows that enable users without programming backgrounds to leverage the integration's capabilities.

Resources

• FME
• Pikterra