6. Reports
After completing any image analysis task in AeroMegh Intelligence drone data analytics software, users can access structured outputs through the Reports module. This chapter explains how to view, interpret, and export different types of detection reports for post-processing, validation, and documentation.
6.1 Introduction to Reports
The Reports module in AeroMegh Intelligence provides users with structured, downloadable summaries of the results obtained from various detection workflows using drone data analytics with ai. These reports transform raw image processing outputs into formatted documents suitable for decision-making, monitoring, compliance, and record-keeping.
Once a detection taskβsuch as Object Detection, Classification, or Change Detectionβis successfully completed on an image, AeroMegh Intelligence drone data analytics software automatically logs the output and makes it accessible through the Reports interface.
Purpose of Reports
Reports allow you to:
- Review detected features in a human-readable format
- Validate detection quality using object-wise and class-wise metrics
- Download results in PDF format for offline use or stakeholder sharing
- Compare changes over time using historical and recent imagery
Types of Reports Supported
AeroMegh Intelligence currently supports the following types of reports:
Report Type | Purpose |
Object Detection | Lists individual objects detected in an image with spatial and class details |
Classification | Summarises the area covered by each class in an image |
Change Detection | Compares two images to identify newly added or disappeared features |
Each report type has a unique layout, analytical depth, and export workflow, which are described in detail in the following sections.
6.2 Accessing Reports
Reports are generated automatically in AeroMegh Intelligence through drone data analytics with ai after a detection task is successfully completed for an image. These reports are accessed from the Images Section within the respective project workspace.
All reportsβwhether Object Detection, Classification, or Change Detectionβare tied to individual images and become visible only if a valid detection task has been executed on that image.
Steps to Access Reports
Follow these steps to view available reports for a specific image:
- Click the Projects tab from the main navigation.
- Open the desired project from the project list.
- Scroll to the Images Section within the project workspace.
- Locate the image row for which the detection task has been completed.
- Click the Reports link on the right side of the image row.
The Reports link is displayed only for images that have at least one completed detection task (Object, Classification, or Change Detection).
Once clicked, this link opens the Reports Page for that specific image.
6.3 Object Detection Report
An Object Detection Report for drone analytics africa displays every object identified within an orthoimage, along with its class, geometry, optional subclass (if classification has been applied), and location data. Reports are viewed in the Monitoring Report interface and can be exported as PDFs for validation, auditing, and documentation.
6.3.1 Opening the Report
- Follow the steps in 2 Accessing Reports to open the Reports table for the desired image.
- Locate the report row corresponding to the Object Detection task for the selected image.
- Click anywhere on that row to open the Object Detection Report.
β This opens the Monitoring Report interface in the same tab, showing one detection card per object.
6.3.2 Report Layout
Once opened, the report displays a structured interface with object-level detection data:
Region | Description |
Header Bar | Displays task name, user name, and detection timestamp |
Preview Image | Static image preview with detection overlays (non-interactive) |
Detection Cards | One card per object, scrollable |
Generate / Download | Top-right control for PDF export |
6.3.3 Exporting PDF
To generate and download the Object Detection Report as a PDF:
- Click the Generate Report button in the top-right corner of the Monitoring Report interface.
- Wait while the button label changes to ..
- Once the report is ready, the button updates to Download.
- Click Download to export the PDF and save it locally.
The exported PDF includes a full list of detected objects with all available fields such as Class Name, Subclass, Area, Perimeter, Volume, and Location.
6.3.4 Detection Fields
Field | Description | Notes |
Thumbnail | Cropped image of the object | UI only (not included in PDF) |
Class Name | Primary label assigned during detection | e.g., Tree, Pit, Vehicle |
Subclass | Classification label (if applied) | May show NA if not classified |
Area (mΒ²) | Surface area of the detected object | Decimal up to 4 places |
Perimeter (m) | 2D perimeter of the detected object | β |
Cut / Fill / Total (mΒ³) | Volume values, if supported by the class | May show NA for non-earthwork classes |
Location (Lat/Lon) | Geographic coordinates of the object | May show NA if geotag is unavailable |
π Fields showing NA indicate that the data is not supported by the detector or image.
6.3.5 Tips
- Use descriptive task names (e.g., Tree_Count_Q1_2025) to improve traceability.
- Check NA fields to verify image layers (DSM/DTM) are present when using volume-based classes.
- Validate large counts using scroll and sampling tools before finalizing exports.
- Store PDFs systematically with project records to comply with quality documentation standards.
6.3.6 Related Sections
Section | Use |
6.4 Classification Report | View subclass values added to this report, if classification was run |
6.7 Change Detection Report | Compare changes between two orthoimages |
3.2.4 Viewing Reports | Return to the report list for this image |
Β
6.4 Classification Report
The Classification Report in drone data analytics software adds sub-class labels to detected objectsβsuch as crop types, material types, or feature categories. This task builds upon Object Detection and does not generate a separate report. Instead, the existing Object Detection Report is updated with subclass information for each object.
6.4.1 How It Works
- Classification is a post-processing step that runs after Object Detection.
- It uses a trained classifier to assign subclass labels to each detected object.
- These subclass values are added to the existing detection reportβnot stored separately.
Object Detection is mandatory before Classification can be run.
6.4.2 Viewing Results
- Follow steps in 2 Accessing Reports to open the Reports table.
- Find the row for the Object Detection task that was enhanced using Classification.
- Click the row to open the updated report in the Monitoring Report interface.
6.4.3 Report Layout
In the Monitoring Report, each object card now includes a Subclass field, in addition to the standard detection fields:
Field | Description |
Class Name | Original object class (e.g., Tree, Pit) |
Subclass | Classification result (e.g., Mango, Neem) |
Area (mΒ²) | Detected object’s area |
Perimeter | Perimeter in meters |
Cut / Fill | Volume values (if supported) |
Location | Coordinates (if available) |
6.4.4 Exporting PDF
- Click Generate in the top-right corner of the Monitoring Report.
- After generation, click Download to save the updated report.
- The PDF will include the Subclass field along with all standard detection metrics.
There is no separate classification reportβthe exported PDF is an enhanced version of the Object Detection Report.
6.4.5 Tips
- Validate detection before classification.
- Choose a relevant classifier for accurate subclass results.
- NA in Subclass may mean low model confidence or unsupported object type.
- Use versioned filenames when exporting PDFs to differentiate detection vs. classification-enhanced reports.
6.4.6 Related Sections
Section | Use |
6.3 Object Detection | Base report updated after classification |
5.3 Classification Task | Steps to run classification after detection |
6.7 Change Detection | View chronological object or area changes |
3.2.4 Viewing Reports | Return to report list for the image |
6.5 Change Detection Report
The Change Detection Report using drone data analytics with ai identifies differences between two orthoimages of the same geographic area, captured on different dates. It highlights objects that have either disappeared (present only in the earlier image) or are newly added (present only in the later image). This report is essential for applications such as infrastructure monitoring, land use analysis, and time-series validation.
The report is opened via the Reports button for the base image and displays a structured comparison of changes, which can be exported as a PDF after generation.
6.5.1 How to Open
- Complete a Change Detection task using the steps in 4 Change Detection Workflow.
- Once the task status shows Completed, go to the Images Section.
- Locate the base image (Image A β the earlier image used during the detection task).
- Click the Reports button on the same row as the base image.
- In the list of reports, identify the row where Type = Change Detection.
6.5.2 Report Layout
The Change Detection Report interface contains four primary regions:
Region | Description |
Header | Displays the report title and timestamp of generation |
Summary Card | Shows names, areas, and timestamps for both the Base and Secondary images |
Results Table | Lists the types of changes detected (Disappeared / Newly Added) with object counts |
Generate / Download Button | Located in the top-right; used to export the report as a PDF |
6.5.3 Exporting PDF
To generate and download the Change Detection Report:
- Click Generate in the top-right corner of the report tab.
- Wait for the report to be processed.
- Once ready, the button label updates to Download.
- Click Download to save the report as a PDF.
The PDF layout mirrors the on-screen interface and includes the header, summary card, and results table.
6.5.4 Results Table Example
Below is an illustrative example of the results table layout:
Image | Date | Change Type | Detection Count |
Base | 01 April 2025 | Disappeared | 1633 |
Secondary | 01 April 2025 | Newly Added | 1509 |
Total | β | β | 3142 |
Counts shown are for reference only. Your results will vary based on image alignment and detector performance.
6.5.5 Tips
- Choose the correct base image: Always select the earlier image as Image A to maintain chronological order.
- Verify alignment: Misaligned orthomosaics may produce inaccurate or inflated change counts.
- Pair with vegetation indices: For agronomic or environmental use cases, combine with 2.6 Index Calculator for a complete change + health assessment.
6.5.6 Related Sections
Section | Use |
5.4 Change Detection Workflow | Describes how to configure and run the Change Detection task |
3.2.4 Viewing Reports | Navigate back to the image-level reports table |
3.4 Detection / Analysing Images | View change overlays in the image viewer interface post-detection |
6.6 Notifications & Task Monitoring
The Notifications Panel in AeroMegh Intelligence for drone analytics africa helps users track the status of processing tasks such as Object Detection, Classification, and Change Detection. It provides real-time updates on task progress and direct access to result views once tasks are complete.
6.6.1 How to Access
- In the top-right corner of the application, click the Notifications icon (π).
- The Notifications Panel slides open, listing the most recent detection and classification events.
6.6.2 Status Indicators
Each entry in the panel displays the task name, associated image, and current status:
Status | Meaning |
Completed | The task has finished processing; results are now available. |
In Progress | The system is still processing the task. |
Pending | The task is queued and will begin shortly. |
Failed | The task could not be completed due to an error. |
6.6.3 Viewing Results from Notifications
- Once a task reaches the Completed status, click the View Results button in the notification entry.
- This redirects you to the appropriate report view for that task type:
- Monitoring Report for Object Detection and Classification
- Change Detection Report Tab for Change Detection
The View Results shortcut eliminates the need to manually navigate through the Projects or Images sections.
6.6.4 Tips
- Use the Notifications Panel to monitor multiple tasks without leaving your current screen.
- Check task status before clicking Reportsβavoid opening incomplete results.
- If a task fails, read the failure reason (e.g., βNo objects detected,β βInvalid image format
6.7 Solving Common Scenarios
Even with a robust detection system, users may occasionally face unexpected outcomes. This section helps you understand whatβs happeningβand get back on track quickly. Instead of focusing on errors, we frame these as common situations with clear solutions.
Scenario | Why It Happens | What You Can Do |
Subclass shows as NA | Classification model didnβt return a valid label | Check if the object type is supported by the selected classifier. Re-run if needed. |
Cut/Fill values are NA | DSM/DTM elevation layers were not uploaded or the class is unsupported | Upload elevation layers before running detection. Ensure classes support volume. |
PDF export not working | Report not generated yet | Click Generate, then wait for the Download button to appear. |
Report missing from the table | Task incomplete, failed, or not yet started | Check Notifications to confirm task status. Retry the task if necessary. |
Change Detection report not showing | Accessed the secondary image (Image B) instead of the base (Image A) | Always open the report from the base image row (Image A). |
Tip: For best results, complete one workflow at a time. Always verify intermediate results (e.g., detection preview) before moving to classification or report export.
6.8 Choosing the Right Report
AeroMegh Intelligence provides multiple types of reportsβeach designed for a specific need. Use the table below to decide which report fits your use case, based on what you want to analyze or export.
Purpose | Recommended Report | Why Use It |
Count and locate individual features | Object Detection Report | Shows every object detected with dimensions, class labels, and map locations |
Break down objects into categories | Classification (within Object Detection) | Adds sub-class insights (e.g., crop type, material) to already detected objects |
Compare whatβs changed between two dates | Change Detection Report | Summarizes whatβs newly added or disappeared in a structured tabular view |
Share or archive structured findings | Any Report (PDF Export) | Provides timestamped, shareable documentation for review, planning, or audit |
Example Use Cases:
- Agronomy: Detect crop zones with Object Detection, classify them into types, then compare season-to-season change using Change Detection.
- Infrastructure Monitoring: Identify and export pits, trenches, or road boundaries. Use volume reports (Cut/Fill) for civil works validation.
- Compliance & Auditing: Export PDF reports after every detection for timestamped evidence and field planning.
