How to Create Point Cloud Cross-Sections: Practical Guidance for On-Site Use

Table of Contents

• Site understanding

• Cross-section setting

• Point cloud selection

• Drafting

• Practical operation

Creating cross-sections from point clouds is not simply a matter of slicing three-dimensional data into lines. It becomes a practically usable cross-section only when you correctly understand the site conditions, clearly define which area is being shown and for what purpose, distinguish usable points from unusable ones, and shape the output so it communicates effectively as a drawing. For field personnel in particular, it is more important—before operating drafting software—to know what to check and in what order to make decisions. Because section drawings are used in many situations—quantity verification, construction planning, as-built verification, negotiation materials, and internal sharing—appearance alone is insufficient. You must check one by one whether the section location is appropriate, whether reference standards are consistent, whether unnecessary points are mixed in, and whether there are any contradictions with actual site conditions. Here, so that field personnel can use it directly for checks, we organize and explain methods for creating point-cloud cross-sections from five perspectives: on-site understanding, section setting, point-cloud selection, drafting, and practical operation.

Site Understanding

The first thing you should do when creating cross-section drawings is not to look at the point cloud data, but to clarify which site will be sectioned and for what purpose. If you start work while this is unclear, you are likely to have to reset section locations or redo unnecessary drafting later. On site, even for cross-sections at the same location, the required presentation changes depending on whether the section is intended to explain things to the client, to compare before-and-after construction, or to check slopes and excavation depths. If the purpose differs, the extent to be viewed, the information to retain, and the dimensions and notes to include on the drawing will also change.

First, clarify the object you want to check on the cross section. How you interpret the point cloud changes depending on whether you want to see the natural ground shape, the pavement surface, the upstand of a structure, or the interface with drainage facilities and gutters. For example, for a slope cross section it is important that the shoulder and the toe of the slope are properly captured; for a road cross section you need continuity up to the shoulder, the gutter, the slope, and the area near the boundary. For a development-site cross section, you need to position the section with an eye toward comparison with the design surface. If you look broadly at the point cloud without deciding on the target object, you will not obtain the necessary cross sections.

Next, what you should confirm is the on-site conditions and the circumstances at the time the point cloud was acquired. A site cross-section is a snapshot of the conditions at the moment of capture. Therefore, if the timing of the photography or measurements differs, vegetation growth, parked vehicles, temporary material storage, puddles, and heavy-equipment tracks can affect the point cloud. Something that looks like terrain may in fact be a temporary obstacle. Conversely, corners or boundaries that should be present may appear faint due to grass or shadows. Creating cross-sections without this background can cause someone unfamiliar with the site to misinterpret the drawings. If you have on-site photographs, plan views, existing drawings, or measurement notes, reviewing them before examining the point cloud will lead to more consistent assessments.

Also, the coordinate reference and the treatment of elevation for point clouds are part of on-site understanding. Because cross-sections rely on vertical relationships, if the coordinate system or datum is ambiguous, the result may look plausible but be unusable in practice. As the site person in charge, you need to confirm early on which reference point the data is based on, whether it is consistent with known control points, and whether there will be any offsets when overlaying point clouds acquired on different days. In particular, when using data acquired multiple times, even if they appear to align in plan before creating sections, subtle differences can emerge in the vertical direction. Cutting sections while leaving these differences unaddressed can lead to misunderstandings such as apparent construction discrepancies or settlement.

Furthermore, it is important to relate the extent required for cross-section drawings to the site’s circulation and the scope of construction. Positions that look good on paper are not necessarily cross-sections that are useful in the field. For example, if heavy equipment access is involved, cross-sections that include bends and sections where the width changes are effective, and if drainage is being considered, cross-sections should include low areas where water tends to accumulate. For excavation management, priority should be given to locations where the planned cross-section and the actual cross-section can be easily compared. Site personnel should, before creating cross-section drawings, be aware who will use the drawings for what decisions and link the on-site usage to the chosen cross-section locations.

If confirmations at this stage are done thoroughly, the uncertainty in later processes is greatly reduced. Creating cross-sections may look like a data-processing task, but in reality the depth of on-site understanding determines quality. If you judge only from point clouds without knowing what is happening in the field, you are likely to pick up unnecessary irregularities or overlook change points that are actually necessary. First, align the site’s purpose, target objects, acquisition conditions, reference standards, and usage scenarios, and solidify the assumptions for making cross-sectional drawings—this is the first step to avoiding mistakes in practice.

Cross-section settings

Once you have understood the site, next you set where to cut the cross-section, in which orientation, and at what width. Section setting is one of the most important steps in determining the quality of cross-sectional drawings. Even with the same point cloud data, shifting the section position by only several tens of centimeters (several inches) can change the visible shape, and if you choose the wrong section direction the changes you want to show may not appear. What field personnel should keep in mind is not to place sections mechanically at equal intervals, but to put them at meaningful positions.

The first thing to consider is the setting of the reference line. For subjects with a clear alignment, such as roads and waterways, the basic approach is to take cross-sections perpendicular to the centerline, street gridline, or the reference line of existing structures. However, on site there are locations—curved sections, widened sections, branches, and transition/merging sections—where a simple perpendicular cross-section does not represent the actual conditions well. In such places, rather than insisting only on the orthogonal relationships shown on the drawings, it is important to choose the orientation that makes the form you want to verify in the field easiest to read. For example, if you force a road-standard perpendicular cross-section where a slope runs diagonally, the slope shoulder or slope toe may be unnaturally cut off, making it difficult to grasp the shape.

Next to consider is the spacing of cross-section locations. While cross-sections at regular intervals are easier to organize, they can miss areas with significant changes. On site, wider spacing may be acceptable in stretches where terrain and structural changes are minor, but at grade change sections, access points, break points at the toe of slope, structure connection points, and transitions between excavation and embankment, additional cross-sections should be added to capture the changes. In other words, combining regularly spaced baseline cross-sections with supplemental cross-sections that capture change points is an effective approach. This makes it easier to both organize the drawings and accurately grasp field conditions.

Setting the cross-section width is also important. If the width is too narrow, the necessary object will not fit; if it is too wide, unrelated information will be included and the cross-section becomes harder to read. Especially when extracting a cross-section from a point cloud, how much thickness you allow in front of and behind the section line to capture points greatly affects the line density and appearance. If the thickness is too thin, there will be too few points and continuity will weaken; if the thickness is too large, points from other positions will mix in and the cross-section shape will become blurred. Field personnel need to choose an adequate thickness based on the object’s size, the required level of representation accuracy, and the point cloud density. The appropriate thickness differs depending on whether you want to see surface continuity, as with pavement or slopes, or clearly see shape breaks, as with curbs or the tops of gutters.

The way the vertical direction is displayed is also part of cross-section settings. In cross-section drawings, the impression changes depending on how the aspect ratio is handled, so excessive exaggeration can lead to a disconnect with on-site perception. You may want to emphasize the vertical direction to reveal slight unevenness, but even then you need to use it with an understanding of the exaggeration factor. Particularly for cross-sections reused in discussion materials or reports, being able to explain the aspect ratio and the reference line settings makes it easier to avoid misunderstandings.

Also, the way section numbers are assigned affects downstream operations. To make it easier for site staff to recheck later, organize them according to rules that are used on site—such as distance from the starting point, survey point, area name, and construction category—so they are practical to use. If it is not immediately clear from a cross-section drawing which location was cut, that will tend to cause rework during field rechecks and internal sharing. It is also important to be aware of the location diagram of the sections and their correspondence on the plan view.

During the cross-section setup stage, it is effective to actually cut a few sections and quickly confirm whether they meet the intended purpose. Rather than creating all cross-sections at once from the start, test at representative locations and, after checking that the slope shoulder is properly revealed, that the pavement gradient can be read, and that no unwanted material is mixed in, finalize the conditions to stabilize accuracy. Although cross-section settings may appear to be easy to change later, redoing positions and widths is surprisingly time-consuming. For that reason, it is important to carefully determine the cross-section positions, orientations, thicknesses, and display settings up front to match the site's objectives.

Point Cloud Selection

Once the cross-section settings are complete, the next necessary step is to appropriately select the points to be used in the cross-section. Point cloud data can include not only the ground surface and structures but also vegetation, vehicles, people, temporary materials, interference noise, and reflection disturbances—points that should not be used directly in the cross-section. The readability and reliability of the cross-section are largely determined by the accuracy of this point-cloud selection. For field personnel, the important task is not simply deleting unwanted points, but discriminating which points represent the actual site conditions and which are temporary or error-induced elements.

The first thing to check is the continuity of the points treated as the ground surface. Because cross-sections are often interpreted as lines, it is important that the flow of the ground surface appears naturally connected. However, in actual point clouds, local irregularities appear in fine detail due to grass, small stones, mud, wheel tracks, and so on. If all of these are reflected directly in the cross-section, even a visually dense section can make it difficult to read the actual terrain or structural trends you want to judge on site. Conversely, if you smooth out too many small irregularities, you risk losing information that should be observed, such as slope failures or disturbances in the construction surface. Therefore, what to retain and what to smooth must be decided according to the intended use of the cross-section.

For example, for cross-sections used for as-built verification, it is necessary to retain some of the actual construction surface undulations. On the other hand, for cross-sections used for plan explanations or conceptual comparisons, it can be clearer to prioritize conveying the overall shape rather than minor noise. In other words, point cloud selection is not governed by a single uniform rule but requires selecting or discarding data according to the purpose. As site personnel, it is important to reconfirm who will view the cross-section and what they will judge, and to strike the right balance between the required accuracy and readability.

The next important point is how to handle the edges and break points of structures. The shoulder of gutters, the upstand of curbs, the top of retaining walls, and slope crests and toes are crucial locations that determine the meaning of a cross-section. If the point cloud is crudely decimated or excessively averaged, these transition points will be rounded off and the as-built shape will be weakened. Conversely, if noise points are mistaken for transition points, they will appear as unnatural bumps or steps. Therefore, for locations on a cross-section that appear to be changes in shape, it is necessary to check—using plan views and photos as well—whether they are truly structural changes or temporary obstructions.

Be aware of contamination by other objects located near the cross-section. For example, if there are fences, signs, vegetation, or temporarily placed pipes near the section line, they may be included as points depending on the cross-section thickness setting. In the sectional view, these can appear to be part of the ground surface or of structures. This is especially true for dense point clouds, where the greater amount of data can make unwanted points look plausible. It is important to distinguish contaminating points by narrowing the display range as needed, using classification information if available, and switching between plan and sectional views while checking.

When surfaces at different elevations are close to each other, the difficulty of cross-section selection increases. For example, along roadside gutters, the boundary between fill and existing ground, terraced (stair-step) development slopes, or the interface between structures and natural ground, a slight change in selection criteria can cause a different surface to be picked up. In such locations, rather than simply retaining more points, it is clearer for the cross-section to decide which surface to show as primary and to prioritize points along that surface. Even when it is necessary to show multiple surfaces at once, if you do not organize them with a clear primary–secondary relationship in mind, the cross-section can become too complex and fail to convey the information.

From the perspective of site personnel using this in practice, it is also important not to trust point-cloud classification results at a glance. Even if neat lines appear as cross-sections, whether they truly represent the site is a different matter. For representative cross-sections, always compare the plan position with photographs, and it is reassuring to double-check especially at change points, low areas, high areas, and edges. Even verifying just a few against on-site records will provide decision criteria for subsequent bulk processing. Point-cloud classification is a modest task, but whether it is done carefully greatly affects the reliability of the cross-sections.

Drafting

Once you have selected the necessary points from the point cloud, the next step is to arrange them so they can be read as cross-sections. What matters here is not simply displaying the data, but organizing the information into drawings that can be used for on-site decision making. A mere collection of points is not sufficient as a cross-section; it must convey what is the ground surface, what is a structure, where the reference is, and which location the section represents. It is easier to understand if you think of the drafting process not as an effort to improve visual aesthetics but as work to enhance the accuracy of communication.

The first thing required is to clarify the references for the section. Section drawings that do not indicate where the section is taken, which direction is being viewed, or which height reference is used are difficult to use on site. It is important to organize the section numbers, station points, relationships to the reference line, left/right orientation, etc., so they correspond to the plan view. If site personnel can grasp the approximate positional relationships from the section drawing alone when reviewing it later, rechecking will be much easier.

Next, consider organizing the shape representation. For cross-sections derived from point clouds, it is possible to show the fine clusters of points as they are, but in meetings and when sharing, line representations that tidy the outline are often easier to understand. However, if you interpolate lines arbitrarily when drawing them, you risk creating forms that do not exist on site. Therefore, while filling in continuity, it is important to correctly preserve change points and break points. In particular, be careful not to make ambiguous lines for locations that serve as origins for dimensions or gradients—such as the slope shoulder, slope toe, curb top edge, gutter bottom, and retaining wall top.

A sense of scale is also important when creating drawings. The appropriate way to present cross-sections depends on the subject. For wide terrain cross-sections it is important to show the overall picture, whereas for narrow structural parts you need a presentation that preserves fine details so they are not lost. If the scale is incorrect, the point cloud data you carefully selected will not be useful. Also, when using vertically exaggerated displays, you must handle the drawings with that in mind. Because cross-sections can circulate on site on their own, consider including dimensions and reference elevations where necessary so that slopes and steps are not judged solely by visual impression.

Moreover, it is essential to decide what to include on the cross sections. More information is not always better; it is important to include elements that suit the purpose of the section. For example, for construction review it is important that the section shows width, elevation differences, clearance from existing structures, slope gradients, and so on, while for as-built verification the differences from the plan and the continuity of the constructed surface are important. Including unnecessary notes or features of little relevance can actually make the drawing harder to read. If the cross sections will be used by site personnel for checks, it is more practical to compose them so that the points to be checked are apparent at a glance.

When organizing cross-section drawings, maintaining a consistent appearance when multiple sections are placed side by side is also important. If the reference positions, display ranges, and annotation rules vary by section, comparison becomes difficult. Especially when using consecutive cross sections to track terrain changes or construction progress, aligning the approach to lateral extents, the vertical datum, and the display ranges makes differences easier to read. Cross-section drawings with a unified appearance also help build trust during internal reviews and client presentations.

Also, at the drawing stage, you should not stop at producing a cross section; you need to adopt an approach of checking back and forth with the plan view. Even if a cross section looks tidy, the object may be misaligned when compared with the plan location. Conversely, something that seems fine on the plan view can suffer from misread heights in the cross section. Iterating among the plan view, cross sections, and site photographs improves the consistency of the drawings. To produce cross sections that site personnel can use as-is, it is important to verify to the end—not just appearance but also that position, orientation, shape, and intended use all match.

Operational Practices

Section drawings are not complete when merely produced; they only become valuable when used on site. For that reason, it is important to organize them with practical operations in mind. A section drawing that is easy for site personnel to use is one that can be referenced immediately when needed, can be used without hesitation during on-site checks and meetings, and still makes the rationale for decisions clear when reviewed later. Here, it is necessary to grasp the operational approach that is useful on site, including how the drawings are handled after creation.

First and foremost, organize cross-section drawings according to their intended use. The required level of detail and the annotations differ for pre-construction checks, interim as-built checks, pre-completion checks, materials for discussion, and internal review documents. If you cram everything into a single set of cross-section drawings, it becomes difficult to reach the information you need. If field personnel will use them directly for verification, organize which sections should be shown for each purpose and make operation easier by clearly indicating cross-section numbers, file names, update dates, and the scope. Especially on sites where measurements are taken multiple times, time-point management is important because comparisons become difficult if you cannot tell which point in time a cross-section represents.

Next, it is important to link cross sections with on-site verification. Always verify representative locations to ensure that the cross sections prepared at the desk match the field conditions, as this will improve operational accuracy. For example, for key decision points such as the shoulder location, the bottom elevation of side ditches, and changes in slope at the pavement edge, treating these together with site photographs and stationing information can compensate for aspects that are easily misread from cross sections alone. At the site, slight discrepancies on the drawings can affect construction decisions, so it is important not to separate office data from field judgment.

Also, in practical operations, making explanations easy to understand for the person you're sharing with is important. Site representatives, construction managers, designers, subcontractors, and owners each look for slightly different information in a section drawing. Rather than giving everyone the same explanation, clarify what you want them to check in this section so meetings run more smoothly. For example, if the goal is to understand the construction scope, prioritize the overall shape; if the goal is quantity review, make reference lines and elevation differences clear; if the goal is discussion, emphasize interfaces with existing structures and points of change. The value of a section drawing lies not in the amount of information it contains but in whether it leads to the necessary decisions.

Update operations should not be overlooked. When point clouds are re-acquired or supplemented, it is desirable to keep them in a state that allows comparison without changing cross-section conditions. If the cross-section position, thickness, or the approach to the reference elevation changes between the previous and current measurements, it becomes unclear whether differences are due to actual site changes or to configuration differences. Therefore, it is useful to record the cross-section rules once they are established so they can be reused for ongoing monitoring and progress checks. As the on-site person in charge, organizing drawing conditions so they can be reproduced from the same perspective next time, rather than relying on memory, will save effort later.

Furthermore, deciding on an order for checks using cross-sections helps stabilize day-to-day operations. First confirm the location on the plan view, next examine height and shape on the cross-section, and finally, when necessary, corroborate with site photos and measurement records — if you establish this sequence, missed checks will be reduced. Conversely, operating in a way that draws conclusions from cross-sections alone makes it difficult to detect positional errors or the intrusion of temporary obstructions. For documents used in the field, it is more effective to make judgments by combining multiple lines of evidence rather than relying on any single source.

Finally, keeping a perspective focused on continuously streamlining the creation of point cloud cross-sections will be useful in practice. If you identify commonly used section locations, structures that require frequent checks, and spots that repeatedly cause problems on site, creating cross-sections will be faster next time. By accumulating, as site-specific check items, the points that must be understood on site, conditions that tend to cause confusion when setting sections, spurious points to watch for when filtering point clouds, and the representations required for drawings, it becomes easier to maintain quality even when personnel change. Point cloud cross-sections are not so much products of advanced 3D processing as practical documents that support on-site decision making. Precisely for that reason, it is most important to keep them in a state where users are not confused, viewers are not misled, and they can be traced later.

In recent years there has also been a growing need to efficiently carry out on-site position checks and additional measurements while linking those results to the verification of point clouds and drawings. In such workflows, having a system that allows you to quickly establish reference positions on site and verify necessary locations on the spot makes it easier to organize the prerequisites for creating cross-sections.

For example, by combining methods like LRTK that make it easy to perform positioning and recording while confirming positions on site, the verification work before and after creating cross-sections from point clouds becomes smoother, and it becomes easier to maintain the connection between the site and the drawings.