How to Create Cross-Section Drawings from Point Clouds and Basic Practical Procedures

Table of Contents

• Why it is necessary to create cross-sections from point clouds

• Preparation

• Setting the reference

• Extraction

• Shaping

• Confirmation

• Summary

Why Cross-Section Drawings Are Required from Point Clouds

The process of creating cross-sections from point clouds is widely used in civil engineering and surveying practice for understanding existing conditions, design review, construction planning, and as-built verification. Cross-sections are highly effective for organizing elevation differences, slope geometries, road cross-sectional profiles, and terrain changes around structures that are difficult to interpret from plan-view drawings alone. Especially when point-cloud data covering a wide area have been captured in a short time, cross-sections can be extracted at required locations, making it easier to confirm site conditions from multiple perspectives.

However, having a point cloud does not necessarily mean you can produce a correct section drawing as-is. Point clouds may contain noise and missing data, and if you work with an ambiguous handling of coordinate systems, the result can look plausible but be unusable in practice. Also, if the method for deciding the section position and width is not standardized, each stakeholder may end up with different section drawings, causing rework during meetings and reviews.

Therefore, when creating section drawings, it is insufficient to merely learn how to crop or extract within the software. It is important to proceed in a sequence of first clarifying the purpose for which the section drawing will be used, establishing common standards, appropriately extracting the necessary sections, formatting them so they are easy to read, and finally verifying the content. By following this flow, it becomes easier to maintain consistent work quality even if personnel change, and the reliability of the deliverables is also improved.

In this article, we explain how to create cross-sections from point clouds by dividing the process into five items: preparation, reference setting, extraction, shaping, and verification. To make the steps easy to follow even for those handling it for the first time, we organize the material as a practical approach while touching on common confusions and mistakes that often occur on site.

Preparation

What matters in the initial stage of creating section drawings is aligning the prerequisites before beginning the work. If this remains unclear and you proceed, you are more likely to need to redo sections or revise drawings later. During the preparation phase, first clarify why you are creating the section drawings. Whether it is to explain the existing conditions, to compare with the design, to assist quantity calculations, or to check for construction management, the required level of accuracy and the method of representation will differ. Because a change in purpose alters the locations of the sections, the number of sections, and the information that should be recorded, it is essential to share the intended use before starting work.

Next to check is the nature of the point cloud data itself. Depending on whether it was measured from the ground, acquired from the air, or captured while moving, the kinds of objects it captures well and poorly will differ. Even if the ground surface is relatively well captured, slope shoulders and the backsides of structures can be missing, and temporary objects such as trees and vehicles may be mixed in. Those responsible for creating cross-sections should not simply accept the point cloud; they need to identify which areas have been sufficiently captured and where there are concerns. If this is overlooked, later stages may find insufficient information along the section lines, requiring supplementary surveying or reprocessing.

In practice, checking the coordinate system is an important matter that should be completed during the preparation phase. If you work while plane rectangular coordinates, arbitrary coordinates, and height references are mixed together, problems such as cross-sections not aligning, not matching existing drawings, and being unable to compare with design values can occur. Confirm whether the point cloud coordinate system, the coordinate system of the existing drawings, and the reference for the drawings to be finally submitted as cross-sections are consistent; if there are inconsistencies, decide in advance where the conversion will be performed. This may seem unremarkable at first glance, but it is a key point that greatly reduces rework in practice.

Also, narrowing down the work area is indispensable in preparation. If you handle the entire point cloud as-is, the data volume becomes large, making display and operation sluggish, and you can be distracted by unnecessary information, which makes judgment difficult. If you first define the cross-section’s target segment, the target structures, the required length, and the range of surrounding information to retain, subsequent extraction and formatting will be more stable. For a road, this might be the segment along the centerline; for earthworks, the range that includes the boundaries of fills and cuts; for a river, the range in which the embankment and slope toe can be checked. It is important to set the target range with awareness of the information you want to read from the cross-section.

Furthermore, having a clear image of the final cross-section drawings from the outset makes it less likely that you will lose your way midway. If you organize in advance items such as which scale to use, whether existing-ground lines alone are sufficient, whether to overlay design lines and control values, how to label cross-section numbers and station point labels, and which direction to standardize as left and right, the post-extraction formatting work will proceed smoothly. Cross-section drawings are not created for their own sake; their purpose is to be readable by others. Therefore, it is important to prepare with the final deliverable in mind.

A common mistake at this stage is assuming that having a point cloud means you can deal with things later. However, in practice, even a slight difference in the cross-section location can make comparisons impossible, and insufficient noise processing alone can change the interpretation of the terrain. Aligning the objective, the area of interest, the coordinate reference, and the expected deliverable during the preparation phase greatly affects the quality of subsequent processes. It is better to think that the success of creating cross sections is often decided by the organization and preparation done before extraction operations, rather than by the extraction itself.

Standard Settings

Once you are ready, the next step is to move on to the reference settings for creating section drawings. By "reference" here we mean the common rules for the entire set of sections—where to place the reference line to cut the sections, at what intervals to create sections, which direction to treat as positive, and which height reference to read from. If these settings are ambiguous, the appearance will differ between sections and continuous comparison will become impossible.

The most important thing is to establish the reference line that will serve as the basis for section positions. For roads, this is the centerline; for rivers, the design alignment or the embankment axis; and for land development, the reference line of the development surface or the locations of management cross-sections—define the reference line according to the subject of the work. Even when the existing shape is complex, treating cross-sectional drawings as comparable deliverables requires placing sections according to a consistent approach based on the reference line. Cutting sections at positions that are visually intuitive is not inherently wrong, but explanatory sections and management cross-sections should be considered separately. For management cross-sections, the basic requirement is that their positioning be reproducible.

The spacing of cross sections is also an element that should be considered in advance in practical work. If the spacing is too wide, changes in shape may be overlooked; if it is too narrow, the amount of work increases and important cross sections can be buried among many others. What matters is not making them uniformly fine, but considering the density according to the degree of shape change. For example, stretches with little terrain variation can be covered by representative cross sections, whereas additional sections are required at slope break points, at interfaces with structures, and where there are large longitudinal changes. Rather than a formal number of sections, it is important to arrange them according to what you want to check.

Section thickness—i.e., the width of points chosen as the cross-section—is also important when setting standards. If the section is taken too thin, there will be too few points and the line will tend to break. Conversely, if it’s taken too thick, points at different depths will mix and the cross-sectional shape will become blurred. In practice, you need to set a width that is neither too thin nor too thick while considering the object’s size, point density, and required accuracy. If the goal is to check the slope of an embankment or cut face, you don’t want to mix in unnecessary depth points; for understanding natural terrain, giving the section some thickness can make continuous shapes easier to read. The appropriate width is not one-size-fits-all and should be chosen depending on the subject.

The handling of elevation references cannot be neglected. In cross-sections, because vertical relationships are read alongside cross-sectional shapes, it is necessary to make clear which elevation reference is used for drafting. When comparing with existing management records or design drawings, a mismatch in elevation reference can make the cross-section line itself look correct while the comparison results are wrong. It is important to confirm here that the current point cloud, the elevation reference, and the values displayed on the cross-section are consistent. The rounding of elevations and the number of displayed decimal places also affect later verification work, so they should be standardized according to project requirements.

Additionally, the left–right orientation and the method of assigning cross-section numbers should be included when setting standards. Because cross-section drawings are often checked side by side across multiple sheets, it places a burden on the reader if the left side alternates between upstream and downstream, or between the start point and the end point, each time. Standardizing cross-section numbers, stationing, direction, and the notation rules on the drawings greatly improves the comprehensibility of the deliverables. It is often thought that such presentation rules can be refined later, but if standards are not decided at the time of extraction, data management becomes complicated and mix-ups are more likely to occur.

It is also important to align stakeholders’ understanding at the standard-setting stage. If the point-cloud team, drafting team, design team, and construction management team do not agree on which cross section will serve as the official reference cross section, they can draw different conclusions from the same data. That is why you should standardize the reference line, cross-section positions, width and height references, and notation rules before proceeding to the next stage. It would not be an exaggeration to say that the quality of cross-section drawings is determined more by consistency of standards than by visual neatness.

Extraction

Once the reference is set, actually extract the required cross-sections from the point cloud. In this step, you extract the points corresponding to the cross-section location and bring them into a state readable as a cross-section. However, the important thing here is not to cut mechanically but to retain only the information that matches the purpose. Point clouds may include not only the ground surface but also vegetation, vehicles, temporary structures, traces of human movement, and other information unnecessary for cross-sectional drawings. It is important to extract while distinguishing between what is necessary and what is not.

First, what you should do is narrow down the area of interest. If you try to cut a cross-section directly from a large-area point cloud, not only will the amount of data be too large and usability deteriorate, but unwanted objects are also likely to be mixed in. Narrow the scope to the target segment and the area around the target structure, and once you have created a state that contains only the information necessary for the cross-section, proceed to extract the section; this stabilizes the workflow. This is not merely a way to reduce data size, but a preprocessing step to preserve the readability of the cross-section.

Another important issue is how to handle noise and unwanted points. If you slice a point cloud into cross sections as-is, unwanted points that happened to be included can become more conspicuous than the ground lines or structures you actually want to show. For example, if roadside grass, branches and leaves on slopes, parked vehicles, or construction machinery remain as-is, the cross-sectional shape becomes disturbed and difficult to interpret. Therefore, you need to adopt an approach to exclude unwanted points at an early stage, either before or soon after cross-section extraction. How far to exclude depends on the objective, but it is important to have the perspective to distinguish between things that should remain as the current condition and those that are temporary.

When setting the section plane, you need to be aware whether to cut sections perpendicular to a reference line or to fix them in a specific direction. For linear features such as roads or rivers, orienting sections perpendicular to the reference line often makes comparison easier, while for inspecting developed land or structures, sections in arbitrary directions can be effective. The important thing is being able to explain why the section is oriented that way. Because section extraction offers a high degree of freedom, it also carries a responsibility to justify it with respect to the objective.

A common difficulty during extraction is balancing cross-section width and point density. If you set the width narrow, the target becomes clearer, but point continuity can be insufficient. If you widen the width, shapes are more likely to connect, but different shapes before and after can mix, producing a thicker, more ambiguous shape than the true cross-section. In practice, rather than deciding once and for all, it is realistic to try several representative cross-sections to find the appropriate width. Especially when you want to accurately capture break points such as slope shoulder, slope toe, side ditch, curb, and road shoulder, you need to adjust while checking that the required shapes are not being lost or smoothed over.

Handling missing data is also an important aspect of the extraction process. In point clouds, gaps can occur in areas with poor line-of-sight or where reflection conditions are severe. When there are gaps in a cross section, you must decide whether it is acceptable to simply connect the lines, whether they should be supplemented with other sources, or whether an on-site recheck is necessary. If interpolation is done lightly here, it can appear as if the shape was actually measured and later cause misunderstandings. Being aware that missing data exist and clearly distinguishing which parts are observed values and which are estimates leads to reliable cross-sectional drawings.

In the extraction phase, it is also important to organize file names and management methods for each cross-section with an eye toward overlaying them with the design cross-sections and existing drawings. Saving multiple cross-sections under the same name or allowing cross-sections with different orientations to coexist will cause confusion during later processing and verification. In practice, simply managing measurement point names, cross-section numbers, orientation, work dates, and so on according to a consistent rule makes it easier to prevent mix-ups. Although the extraction step may appear simple, proceeding while considering its connections to downstream processes can greatly affect overall efficiency and quality.

Plastic surgery

It is often not possible to use extracted cross-section points as deliverables as-is. The shaping process arranges the sequence of points into an easy-to-read cross-sectional drawing, converts them into linework where necessary, and brings them to a state suitable for comparison and explanation. This process requires not only improving visual appearance but also appropriately representing the current shape. If you tidy it up too much, it will diverge from the actual condition; if you don't tidy it up enough, it will become hard to read. Striking the right balance between the two is the key in practice.

First, you should consider which points to adopt as the existing-condition line. Point-cloud cross-sections can exhibit many small irregularities, and if you trace every one of them exactly with a line, the terrain’s or structure’s characteristics can actually become harder to discern. Conversely, if you oversimplify, important break points such as slope shoulders, road shoulders, gutter bottoms, and the edges of structures may be lost. When smoothing, it is important to level out fine details while preserving the meaningful aspects of the shape. How much of the existing features should be represented must be judged according to the intended use of the cross-section.

For example, if a cross-section is used for quantity calculations or design comparisons, the positions and heights of the break points should be handled carefully. Rounding too much to improve appearance can affect the interpretation of cross-sectional area and slope. Conversely, for presentation materials it may be better to prioritize suppressing minor noise so the overall shape is conveyed. In other words, shaping should be regarded not as a task of making things look neat, but as an adjustment of the amount of information to suit the purpose.

When preparing a cross-section drawing, consider how to include reference lines and auxiliary information. For example, when comparing with the planned section, the layout needs to make the differences from the existing condition line readable, and you should depict elevations, offsets, and the positions of key points so they are clear. However, adding too much information can make the drawing cluttered and, paradoxically, harder to read. Rather than cramming a lot into a single cross-section, clarifying what you want the viewer to see and organizing the drawing accordingly makes it more usable in practice.

In the shaping process, you should also be mindful of continuity with adjacent cross-sections. A single cross-section may look natural on its own, but when compared with the preceding or following sections, the height can change abruptly or the relative positions of breakpoints may appear unnatural. Such inconsistencies arise from variations in extraction width, swapped orientations, the inclusion of unnecessary points, or tendencies in linearization during shaping. Therefore, rather than working on each cross-section in isolation, it is important to arrange them side by side and refine them while checking for any irregularities.

Also, when section drawings are to be treated as final drawings, organizing layers and notations is indispensable. Properly separating existing-condition lines, planned lines, auxiliary lines, and notes makes revisions and sharing easier. This is not merely a drafting convention but a measure to improve the maintainability of the deliverables. Considering that another person may need to modify them later, it is very important to make clear what each line represents. Because urgent revisions often occur on site, keeping the drawings in an easily manageable state during the formatting stage makes subsequent work considerably easier.

One thing to be careful about during the refinement process is not to lose the connection to the original point cloud data. Once it has been linearized, neat-looking lines can easily take on a life of their own. However, the reliability of cross-section drawings is supported by the original observation data. Keeping the ability to trace which line was created from which point cloud and where interpolation or corrections were applied makes the deliverable stronger in checks and explanations. Refinement is both a visual finishing step and a process that enhances the explainability of cross-sections.

Confirmation

The final check carried out at the end of creating cross-section drawings is not a mere review but a process to determine whether the deliverable can be used. At this stage, you need to carefully inspect not only for typographical errors and formatting but also whether the cross-section locations are correct, whether the interpretation of shapes is reasonable, and whether there are any contradictions with the established reference settings. Because once cross-section drawings are distributed, design and construction decisions may proceed based on them, the final check is highly important.

The first thing to do is to verify it against the original point cloud. Even if the smoothed cross-section lines look fine, returning to the original point cloud may reveal that extraneous points were picked up or important break points were omitted. It is important to compare with the source data to confirm that key points that define the shape—such as the slope shoulder and toe, edges of structures, and pavement edges—are properly reflected. Having someone other than the operator review the work can also increase the number of issues noticed.

Next, check for consistency with the reference. Verify whether the cross-section positions are as specified, whether the cross-section orientations are uniform, whether there are any offsets in the height datum, and whether there are errors in how measurement points or cross-section numbers are assigned. Especially when handling multiple cross-sections, it is effective to check them in a list, because similar names can be confused and orientations can be reversed. A cross-section that appears correct on its own may reveal abnormalities when viewed in the context of the entire sequence.

Furthermore, readability as a cross-sectional drawing is also subject to review. We check whether necessary information is missing or, conversely, whether there is so much information that the main point becomes difficult to convey. Problems such as existing-condition lines overlapping comparison lines and becoming hard to distinguish, an inappropriate scale for the section, or difficulty reading the positional relationships of key points will make a deliverable hard to use in practice even if the content is correct. A cross-sectional drawing must not only be accurate but also make it easy for stakeholders to arrive at the same interpretation.

Don't forget to check the continuity between cross-sections. By confirming whether shape changes are too abrupt compared with the preceding and following sections and whether there are inconsistencies in the terrain's connectivity, you can more easily detect extraction and shaping errors. If only a particular cross-section shows an unnatural protrusion or depression, you should not just correct that section alone but trace the cause back to the original point cloud, the extraction width, processing of unwanted points, and the lineization method. Superficial fixes alone may leave the same problem hidden in other sections.

In practice, it is also necessary to check with the recipient or the party with whom the material will be shared in mind. Whether the document is handed to the design team, the construction team, or used for explanations to the client will change the level of detail required in the notation. What is sufficient for internal review may lack notes when shared externally. Conversely, expressions that are too detailed for internal work can make explanatory materials harder to understand. Making final adjustments with the assumed use of the deliverable in mind increases the practical usefulness of section drawings.

It is also useful to briefly organize and record the conditions used to create cross-section drawings. Make sure it can be traced which point cloud was used, which reference line the sections were based on, how the section width was set, and what judgments were made during shaping; doing so makes corrections and reuse easier. In practice, cross-section drawings are often re-referenced at later stages, and if the working conditions are unknown then reproducibility is lost. The verification process not only preserves current quality but also serves to prepare for future modifications.

Summary

The task of creating cross-sectional drawings from point clouds may seem simple if you only look at the operational steps, but in practice it only becomes a usable deliverable by carefully following the sequence of preparation, standard-setting, extraction, shaping, and verification. During preparation, clarify the purpose and target scope, the coordinate reference system, and the expected deliverable image; during standard-setting, unify the cross-section position and orientation as well as the width and height reference criteria. From there, consciously extract cross-sections from the point cloud with the necessary information in mind, shape them to the amount of information appropriate for the intended use, and finally verify them against the source data and overall consistency.

Keeping this flow in mind improves quality not only in the appearance of cross-sections but also in how easy they are to explain, compare, and reuse. Point clouds contain a great deal of information, but as they are they can be hard to interpret. That is precisely why, when converting them into cross-sections, it is important to be clear about what to keep, what to omit, and by what criteria to present them.

In situations where you want to perform supplementary measurements and verify reference points on site, an environment that can quickly handle high-precision positioning information is useful. For example, using an iPhone-mounted GNSS high-precision positioning device such as LRTK makes it easier to link point cloud data with on-site verification, and can also improve the accuracy of reference checks and supplementation when creating cross-section drawings. Rather than completing the work with point cloud processing alone, adopting a workflow that combines it with on-site reference verification is the shortcut to consistently producing cross-section drawings that are usable in practice.