Specialists of KB "Panorama" have prepared a video tutorial on automated updating of digital maps using the Panorama Vision service within GIS "Panorama". The training material demonstrates a comparison of objects from the original digital map with objects recognized on a new satellite or aerial photograph.

The video material introduces users to the functionality of the "Search for unregistered objects in an image" task and shows the complete data processing cycle – from selecting the source image and analysis area to generating an updated vector map. The selected image area is sent to the Panorama Vision service, where a neural network model performs object recognition and vectorization.

The recognition of objects of the "Residential buildings" class is considered as an example. The neural network model generates building contours, which are then automatically compared with objects from the original digital map. Based on the analysis results, objects are color-coded depending on the identified changes and the degree of correspondence with the reference data. This allows for quickly identifying buildings missing from the map, as well as objects with changed contours. The video tutorial demonstrates the automatic placement of labels on the map and the application of additional processing parameters to improve the accuracy of the generated geometry. In particular, the setting of the maximum distance between points of compared objects is discussed. The user can adjust processing parameters considering the terrain characteristics, the quality of the source image, and the requirements for the result.

The presented solutions can be applied in the creation and updating of digital city plans, execution of urban planning and cadastral works, territory monitoring, assessment of emergency consequences, as well as in solving defense and security tasks. Using the service reduces labor costs for image processing, decreases the amount of manual data entry, and increases the completeness and reliability of cartographic information. The application of Panorama Vision as part of GIS "Panorama" enables a transition from traditional manual vectorization to intelligent spatial data processing. Objects recognized in the image are automatically converted into a digital terrain model, which can be supplemented with semantic characteristics and address information.

The video material is available in the "Video tutorials" section.

KB "Panorama" has developed GIS "Operator" version 15.5.5. In the new version, tools for automatic control of specialist actions when placing objects on the map and entering attributes have been improved. Settings for the design of object characteristic labels have been expanded. Capabilities for updating semantic characteristics and configuring object creation parameters have been enhanced.

The "Map Editor" task has been improved. In the modes for creating and editing object contours, the ability for automatic metric control for self-intersection and inclusion of sub-objects has been added. The control is performed when writing a linear or area object to the map. If errors are detected, a pop-up message appears, a transition to the point with the first error is performed, and a marker is placed. The complete list of errors is written to a log, which opens in the "View Control Results" mode. By checking the "Check metric on save" checkbox in the "Vector Map Editor Parameters" dialog, automatic metric control is activated in object creation and editing modes. This allows for metric correction more quickly, reducing the need to run control for the entire map.

The "Create Object" mode has been updated. Added the ability to check the quality of object contour digitization. In the "Vector Map Editor Parameters" dialog, the "Check new segment length" checkbox enables highlighting of a drawn segment if it does not meet the vectorization parameters. If the segment length is less than the "Filter level" field value, it is highlighted with a red semi-transparent line. If the segment length is greater than the "Tracing step" field value, the portion of the line exceeding the tracing step is highlighted with a purple semi-transparent line. Segment highlighting helps visually determine the recommended position of the next point – until purple highlighting appears. This method of drawing allows reducing the number of unnecessary points, as well as ensuring a given point density during manual vectorization of terrain objects. The updated mode allows adding, deleting, and changing the order of columns with object information. Managing the table composition is done using the context menu, which lists object attributes (display scale, presence of 3D view, digitization direction, meta-object). Multi-level sorting by object attributes has been added using the Shift key combined with clicking on a column header. The order and type of sorting are indicated in the headers of sorted columns, for example: "[1] Code", "[2] Name". Filtering of classifier objects is performed based on the displayed table columns. Available localizations, the layer tree, and classes are automatically filtered based on the selected objects. Layer or class search has been added with a dropdown list of selected values. For objects having a 3D view, three-dimensional display has been added.

The "Update Object Semantics" dialog, which replaces the values of semantic characteristics of selected objects, has been improved. The new version allows adjusting semantic characteristics based on entered values or using a dictionary (in CSV format). The dictionary string format is: "original value";"new value". It is recommended to use ";" or a tab as the delimiter. Double quotes are recommended but not required. The search for the original value is performed in modes: Exact match, Partial match, Match from beginning. When selecting the "Partial match" or "Match from beginning" property, a "Replace partially" checkbox is available. When partial replacement is enabled, the search and replacement of a specific word or word combination within the semantic value is performed. If the text of the original semantic string contains words specified in several dictionary lines, that string will be processed several times – until all found words have been processed.

Added the ability to select and update values for several codes of the same type of semantics (numeric, text, angular, and other types). Dictionary-based semantic update is performed only for text semantics. By default, the semantics list contains all types of semantic characteristics. In this case, only one characteristic can be updated. To update multiple semantic characteristics, you need to select a type and highlight several lines in the semantics list. Selecting multiple semantics is done with the left mouse button in the semantics list while holding the Shift or Ctrl key. The task is called from the "Map Editor - Edit Semantics" panel.

The display of formatted labels has been improved. Text editing is performed in the map editor's "Edit Label Text" mode or in the "Object Properties" dialog under the "Metric" tab. Usually, the field for entering label text contains the text itself. However, this text is not linked to the specific semantics of the object. Therefore, after editing the object's semantics, the label text on the map may differ from the text in the semantics. To eliminate contradictions between the text in the metric and the semantics (in the "Metric" tab), a reference to the semantic value is entered into the text entry field, for example, to semantics with code 9 (Proper name): "#9". In the new version of the program, extended formatting has been added – after the semantic code of a numeric type, the label output format "%0N.n" or "%N.n" can be specified, where "N" is the total number of characters of the number, including "." and "-" signs. The "0" character after the "%" sign indicates left padding with zeros if the integer part of the number occupies fewer characters than specified by parameter "N". This notation is typically used for displaying angle values – format "#321%03": for semantics with code "321" with a value of "7", the label "007" will be displayed on the map, and with a value of "0" – "000". The parameter "n" indicates the number of digits after the decimal point for a real number. If parameter "n" is not specified, the value is rounded to an integer. The new formatting rules can be applied to labels and vector signs that have labels referencing semantics.

When importing vector terrain data in KML format, support has been added for links from the KML file to external KML files, which can be packaged into a common KMZ archive or specified as links (URLs) to external resources.

The new version of the program is available for download in the Downloads section.

Specialists of KB "Panorama" have prepared an educational video for GIS professionals, showing how to combine the Python language, the Cursor development environment, and the capabilities of artificial intelligence to create custom spatial data processing scripts in GIS "Panorama". The lesson is dedicated to hydrological analysis with automated placement of cross-section points along the river thalweg.

To construct a breakthrough wave graph in the event of a dam failure at a hydraulic structure, calculations of wave parameters at cross-section points are performed. Determining the location of these cross-section points is currently done using an expert method with visual analysis of the map. Using standard GIS tools, such operations amount to a chain of repetitive actions: measuring lengths and heights, placing points, calculating river parameters at the cross-section, and transferring values to the cross-section attributes. Automating this stage significantly reduces the time needed to prepare initial data for calculations and eliminates manual entry errors. The video tutorial proposes the following approach: a specialist formulates a task as a detailed prompt, and an AI agent in Cursor, relying on the open repository panorama-sdk, prepares the basis of a Python script using MAPAPI calls. The resulting script is connected to GIS "Panorama" and executed on a prepared project that includes the hydraulic structure dam, the river thalweg downstream of the dam, and a height matrix in the potential flood zone.

The analyzed example demonstrates the advantage of automating script writing specifically for spatial analysis. The script places cross-section points along the thalweg, starting from the hydraulic structure dam with a specified step and at a specified distance. For each point, the width of the normal flow is calculated; using the height matrix, elevations of equal heights are determined to the left and right of the thalweg, multiples of the contour interval; the shape of the river floodplain and the distance between height marks are calculated, and the floodplain width is limited to an acceptable value if necessary. The operator only sets the step between cross-section points and the contour interval step in a dialog – the rest is performed according to uniform rules, making the calculations reproducible and relieving the GIS specialist of the bulk of the routine work. Instead of manually assembling code, one can more quickly proceed to verifying the analysis logic, geometric correctness, and compliance of the result with project requirements. The proposed methodology can serve as a convenient starting point for mastering Python in general and the MAPAPI interface for applied spatial data processing.

The results of determining cross-section points are used in the "Simulation of a Breakthrough Wave during the Destruction of a Hydraulic Structure" mode within the "Hydrological Tasks Complex". The complex is oriented towards supporting the full cycle of work with hydrological information within the GIS "Panorama" technology suite and helps consolidate standard calculations and data preparation in a single application environment. Combined with the ability to develop and connect custom Python scripts, including the use of AI in their preparation, such a complex expands the specialist's toolkit: from standardized hydrological operations to custom algorithms for a specific water body or territory.

You can view the video tutorial in the "Video Tutorials" section.

Leading experts in the field of land and property relations from 50 constituent entities of Russia became participants in the IV Annual Interregional Land Forum "Mastor". The event took place in Saransk on May 28 and 29, with the government of the Republic of Mordovia acting as the organizer. In addition to representatives of federal and regional authorities, business, and science from Russia, the key industry event was attended by guests from Belarus, Oman, Bahrain, Egypt, and Azerbaijan. Forum participants discussed current issues of land and property relations, best practices, and legislative initiatives in the field of land use.

Representatives of KB "Panorama" presented a report on the topic "Spatial Data as a Tool for Territory Capitalization", emphasizing that the land and property sector should help develop the region and attract investment. It was noted that in modern conditions, a territory competes not only based on natural resources, transport and engineering infrastructure but also on the quality of its digital description. An investor makes decisions based on comprehensive information about land plots, real estate objects, engineering networks, transport accessibility, urban planning constraints, and development prospects. The Federal Fund of Spatial Data, the Unified State Register of Real Estate (EGRN), and the National System of Spatial Data form a unified information base for territory management, development planning, and increasing the investment attractiveness of regions.

Special attention in the report was given to the application of artificial intelligence technologies, computer vision, and spatial analytics. KB "Panorama" solutions, including Professional GIS "Panorama", Panorama Vision, Panorama Photo, and geoinformation systems at the federal, regional, and municipal levels, allow automating the creation and updating of maps, identifying unregistered real estate objects, analyzing actual land use, assessing transport accessibility, and identifying promising sites for business.

The integration of GIS with predictive models and analytical services turns spatial data into a practical tool for making managerial decisions. As a result, high-quality geodata become the foundation of the territory's digital economy, help increase the transparency of property relations, increase budget revenues, and accelerate the implementation of investment projects.

Photo materials were provided by the event organizers.

Specialists of KB "Panorama" have prepared an educational video on the application of artificial intelligence, the Python language, and GIS "Panorama" for automating spatial data analysis. The video tutorial uses a practical example to show the automation of designing an AF maneuver according to ARINC Specification 424-21. ARINC is one of the key families of standards used in describing and exchanging aeronautical information for aviation systems. This example examines the construction of an arc by center, end point, and start direction angle. This demonstrates the connection between the formal requirements of the standard, geometric calculations, and their implementation in a digital map.

The video tutorial demonstrates how modern AI tools help create custom scripts for GIS "Panorama". The user formulates a task in natural language, and the agent analyzes materials from the open panorama-sdk repository and prepares a Python script using MAPAPI capabilities. This approach lowers the barrier to entry into programming, accelerates the learning of applied automation, and helps the GIS specialist focus on the logic of spatial analysis, result control, and the requirements of a specific project.

The lesson shows the complete workflow: preparing a detailed prompt, generating the script, connecting it to GIS "Panorama", selecting source objects on the map, entering the angle parameter, and checking the constructed arc against the reference result. This example illustrates that automating script writing is useful not only for reducing manual operations but also for increasing the reproducibility of calculations, especially when geometric constructions need to be performed repeatedly according to specified rules.

The practical value of this approach is particularly noticeable when working with aeronautical data, where precision, formalized object description, and compliance with industry requirements are important. The Aeronautical Information Documentation Complex is designed for maintaining, analyzing, and generating aeronautical data in digital form. Combined with GIS "Panorama", Python, and artificial intelligence tools, it allows specialists to more effectively solve tasks related to preparing aeronautical information, verifying spatial constructions, and developing their own application-specific data processing scenarios.

You can view the new video material in the "Video Tutorials" section.