6 Causes and Fixes for Being Unable to Extract Cross Sections After Loading an LAS File

Table of Contents

• What to check first when you cannot extract cross-sections after loading LAS

• Cause 1: Coordinate or elevation assumptions are incorrect

• Cause 2: The position or orientation of the cross-section line is misaligned

• Cause 3: The extraction width setting is not appropriate

• Cause 4: Too many irrelevant points are obscuring the required cross-sections

• Cause 5: Sections are being cut without checking point cloud density or missing data

• Cause 6: Extraction is completed without the prerequisites for generating cross-section drawings

• Practical steps to stabilize LAS cross-section extraction

• Summary

What to clarify first when you cannot extract cross-sections after loading an LAS file

When you can load an LAS but cannot properly extract section drawings, many practitioners tend to assume the extraction function itself is at fault. However, in reality, many of the causes of failure in section extraction are due to insufficient checks after loading or proceeding with work while the purpose of the section drawings remains unclear. The point cloud may display yet the desired section doesn't appear; a section may appear but not in the shape you want; or something that looks like a section exists but cannot be used as a drawing. Such situations often occur because the conditions were not properly set before pressing the extraction button.

Cross-section extraction is not simply the task of viewing a point cloud from the side. It requires decisions about where to cut, in which direction to cut, how wide a band of points to collect, and what to keep or exclude. Moreover, the optimal cross-section conditions change depending on whether it is a terrain cross-section, slope inspection, or verification of the shape around structures. In other words, the workflow from LAS import to cross-section extraction is a continuous sequence of multiple steps: display, inspection, organization, trial extraction, parameter adjustment, and diagramming.

Additionally, although LAS data contain a very large amount of information, they are not in a state suitable for cross-sectional drawings as-is. They can include not only ground surface points but also vegetation, temporary structures, materials, surrounding obstructions, stray points, and thickness caused by overlapping scans. Even if these do not appear to be major problems in plan view, unnecessary points can become conspicuous in cross-sections and obscure the contours you actually want to see. Therefore, when extracting cross-sections, how you organize the point cloud can greatly affect the results.

Furthermore, what is needed in practice is not a cross section that only needs to be seen once, but a cross section that can be reproduced, compared, and used as a drawing. A cross section cut on an ad hoc basis may look plausible at first glance, but you may not be able to reproduce the same conditions at another location, and orientations or reference points may not align when multiple cross sections are placed side by side. In other words, behind the complaint that cross-section extraction cannot be done, there is often not mere unfamiliarity with the操作 but the deeper problem that work standards have not been established.

Below, we divide six common causes that tend to occur when cross-sections cannot be extracted after importing LAS, and we organize concrete, practical countermeasures for each. Simply isolating the causes makes it much clearer what needs to be fixed to stabilize cross-section extraction.

Cause 1: Assumptions about coordinates and heights do not match

As a cause of being unable to properly extract cross-sections after loading an LAS file, the first thing to suspect is the assumptions about coordinates and elevation. It’s easy to feel reassured when you can open the file, but even if it appears to be displayed correctly, the interpretation of the coordinate system or the elevation assumptions may actually be offset. In this state, drawing a section line may not cut the desired location, or the elevation representation may look unnatural, making it difficult for the cross-section to be valid.

For example, even if an object appears close by in plan view, if the coordinate units or the origin differ it may not actually align with other data or reference lines. In particular, when you want to generate cross sections while reconciling multiple survey lines, design data, and existing drawings, planar discrepancies will appear directly as shifts in the cross-section positions. Also, differences in the interpretation of Z can make vertical relationships look unnatural when converted to cross sections, causing misinterpretation of the ground surface or the shapes of structures.

The troublesome thing about this issue is that it’s hard to notice from plan view alone. Even if elements seem to fit plausibly when viewed from above, the moment you look from the side their heights may be extreme or their relationship to the reference plane may not match. When sectional drawings do not appear, or even when they do but look odd, you should first question the assumptions about coordinates and heights before touching the extraction conditions.

As a countermeasure, immediately after loading it is important to check the point cloud not only in plan view but also from the sides and oblique angles, and to verify whether its positional relationships align with known reference points and existing drawings. For the vertical direction, you should also compare against known top elevations, road surface heights, and representative heights of structures to see if there are any extreme discrepancies. If there are problems at this stage, it is quicker to fix the loading settings or the way coordinates are handled first, rather than adjusting the section extraction conditions.

Also, when overlaying multiple datasets, it is important to clarify which dataset will be used as the reference. Even if a feature is visible in the LAS alone, if it is offset from the design baseline or other current-condition data, the positioning of the section line itself will become unstable. Extracting cross-sectional drawings can easily appear to be a matter of how you slice them, but if the prior alignment is not consistent, no amount of adjustment will make them settle.

In practice, when a cross-section drawing doesn’t appear people tend to immediately suspect the extraction function, but misinterpretations of coordinates and heights are very common initial causes. Simply checking this first can significantly reduce the amount of trial and error later.

Cause 2: The position and direction of the section line are misaligned

Among the frustrations of being unable to extract cross-sections after loading LAS files, a common case is that the position and orientation of the section line do not match the intended purpose. If you can open the file and see the point cloud but the extracted section does not take the shape you want, you should strongly suspect this cause. Because the appearance of a cross-section changes greatly depending on where and in which direction you cut it, even a slight misalignment in position or direction can make the section unusable in practice.

If you want to view cross sections of a road or walkway, sections cut perpendicular to the centerline or reference line are usually easier to read. If you want to see the shape of a slope, you need to cut in the direction that most clearly shows the slope's variations. If you want to check the edges or faces of a structure, cutting in a direction as directly perpendicular to that face as possible makes it easier to capture the outline. In other words, the orientation of the section line should be determined by its relationship to the object; lines drawn at random rarely work well.

The position of the section line is just as important. If it doesn't pass through the center of the area you want to see, a section may be produced but the crucial shape won't appear clearly. For example, if you want to observe changes in the slope shoulder but cut slightly off, the desired break point becomes ambiguous. Even for structural sections, if you miss the center when cutting, it becomes difficult to understand the left-right balance and the detailing at the ends. When a section drawing is hard to read, before suspecting extraction width or poor organization, you should first check whether the cutting position itself is correct.

An effective measure is to first clarify the purpose of the cross-sections and establish a reference line that fits that purpose. For roads, use the centerline or alignment; for slopes, adopt a normal-direction approach; and for structures, be mindful of the orthogonality to the target surface—avoid cutting sections one by one on an ad hoc basis. If it’s difficult to determine the perfect position from the start, try several provisional cross-sections at representative positions to identify which orientation and position best meet the purpose, which will make subsequent extraction more stable.

Also, when taking multiple cross sections, it is not enough for just one to be easy to view. You also need to consider comparison between sections and reproducibility. For that reason, it is important to standardize which reference line you use and the rules for placing the sections. If you cut the first one by feel, the conditions for the second and third will not match, making them difficult to compare when arranged side by side.

The position and orientation of the section line are the fundamental conditions that determine the extraction results. If a section view does not appear, or if it does appear but does not show the desired shape, you should first consider the possibility that the position and orientation of the cut itself do not match the intended purpose, rather than assuming there is an issue with the extraction function.

Cause 3: The extraction width setting is not appropriate

When cross-section extraction fails, even if the section line itself is correct, the failure can be caused by an inappropriate extraction-width setting. This is a very common cause, yet an aspect that is easily overlooked. People tend to imagine a section as being cut by a single line, but in practice points are usually sampled from a fixed width around that line to generate the section, so this width setting greatly affects the results.

If the extraction width is too large, a large number of unwanted points located in front of or behind the cross-section will be included. When you want to view the ground surface, points from background vegetation or structures may mix in and make lines appear thicker, and when you want to view a wall surface, interfering objects before or after can enter and make the outline unreadable. Even if it looks like a small width in plan view, in the cross-section all information within that width is displayed overlapped, making it harder to see than you might expect.

On the other hand, if the extraction width is too narrow, it will not capture enough of the required points. While this may be acceptable at sites with high point density, in places where density falls off slightly or where the target object is thin, a narrow width alone can cause cross-sections to break up, and the points become sparse and hard to recognize as lines. As a result, even though cross-sectional views are produced, the essential ground surface and contours become fragmented, making them difficult to convert into drawings.

This problem tends to occur because people assume that once the extraction width is decided, the same conditions will be fine everywhere. However, in practice the optimal width changes depending on the size of the target object, point density, the amount of noise on site, and the desired accuracy of the cross-section. The appropriate width differs between road cross-sections and structural cross-sections, and even within the same site the way points appear varies by location. For that reason, it is more stable to treat extraction width as something to be adjusted according to the purpose rather than as a fixed value.

As a countermeasure, instead of immediately performing a full extraction of all cross-sections, it is effective to try several provisional cross-sections at representative locations. Compare wider and narrower widths to check how much easier the ground surface and contours become to read. Based on that, decide how much thickness to allow for the target, which makes it easier to settle on the conditions. Furthermore, for cross-sections with different objectives, it is important to consider different widths; even simply not using the same width for terrain purposes and for structure inspection will considerably improve readability.

Also, if something feels off after extraction, rather than forcing line cleanup in later stages, it's more efficient to go back to the width settings and readjust. Manually fixing cross-sections that contain too many unnecessary points afterward becomes increasingly burdensome as the number of sections increases. Setting the width appropriately from the start will ultimately be faster and more stable.

The extraction width may seem like a minor setting, but it is an important factor that determines the quality of section views. If a section is not visible, or is visible but unusable, be sure to check not only the position of the line but also whether the width suits the target.

Cause 4: Too many unnecessary points are burying the necessary cross-sections

When you feel you cannot extract a cross-section after loading LAS, it may not be that the section isn’t generated, but that the required section is simply buried and hidden by unwanted points. This is a problem specific to point clouds: irrelevant objects that weren’t much of an issue in plan view can suddenly interfere when viewed as a cross-section. It tends to occur in both terrain and structure cross-sections, resulting in symptoms such as cross-section drawings being hard to read, lines appearing thick and unreadable, and being unable to pick out only the desired contours.

The most typical examples are vegetation and plants. When you want to see the ground surface, if many points from the tops of grass or from branches and leaves appear in the section, the true ground surface becomes obscured. The same applies to temporary structures and materials: even items that were placed there temporarily are recorded in the point cloud, so if they appear in the section they interfere with reading the ground or wall surfaces. Around structures, fences, foreground equipment, other structures, and surrounding background points can mix in, making it difficult to discern the target outline.

If you force yourself to create section drawings in this condition, you'll lose a lot of time on post-extraction drafting. Even if you try to pick only the necessary lines from the point-cloud cross-section, when unwanted points are dominant it becomes difficult to discern the true shape itself. Therefore, the basic countermeasure is to reduce unwanted points before extraction rather than to correct them after extraction.

If classification information is available, it can be a great help at this stage. If a ground-surface class exists, narrow the selection to the ground and reduce vegetation and other points, and the readability of cross-sections will improve significantly. However, classification is not万能. Depending on site conditions, points close to the ground surface may be labeled as vegetation, or conversely some vegetation points may be mixed into the ground-surface class. Therefore, use classification results as an aid, but ultimately inspect the target cross-section and decide whether unnecessary points have been reduced.

An effective countermeasure is to check the area around the target in both plan view and side view before extracting cross-sections, and to determine from which direction unwanted points are entering. Even if it looks fine from above, many points may remain at higher positions when viewed from the side. Because these points can all appear overlapped in a cross-section, it is important not to judge based on the plan view alone.

It is also necessary to change the organization strategy depending on the subject. For terrain cross-sections, prioritize organizing the ground surface; for structural cross-sections, adopt an approach that reduces interfering points unrelated to the target structure. Rather than applying the same organization rules to all cross-sections, in practice it is more efficient to reduce unnecessary points based on what you want to see in the cross-section.

A situation in which there are many unnecessary elements and the required section drawings are buried is not a lack of functionality but a lack of organization. When you feel you cannot extract section drawings, it is important, before repeating the extraction operation, to determine what is obscuring the section drawings.

Cause 5 Cutting without checking point cloud density or missing data

Even when the section line and extraction width both look correct, if the cross-section is interrupted, does not form a continuous line, or only the necessary areas have sparse points, the cause can be inadequate checking of point cloud density or missing data. This is a very practical cause and is easily mistaken for a problem with the extraction parameters. In practice, there are many cases where people waste time trying to fix things only by adjusting the parameters, even though there simply aren't enough points.

Although a point cloud may appear to uniformly record an entire site, the density and quality vary by location. Surfaces angled relative to the measurement direction, shaded areas, around thin members, highly reflective surfaces, and the backside of obstructions often have fewer or missing points. Even if a planar view appears sufficient, a cross-section can reveal gaps in specific areas, making it difficult to discern the required shape. This is not a failure of the extraction function but rather a condition of the original data.

A common mistake at this stage is trying to compensate for too few points by simply widening the extraction width. It's true that widening the width will increase the number of points, but it also makes it easier for unnecessary points to get mixed in, and ultimately the cross-sectional view becomes thicker and harder to read. Trying to make up for a lack of points solely by increasing the width often creates other problems.

As a countermeasure, it is important to first understand the bias in point-cloud density and the locations of missing data before extracting cross-sections. Check not only the plan view but also side and oblique viewpoints to see which parts are sparse and where points are captured continuously. Based on that, consider measures such as slightly shifting the cross-section line, changing the cross-section direction, or taking a separate representative position, as these may be more effective than forcing width adjustments.

Also, it is important to recognize missing data as missing. In practice, forcing a line where there are no points mixes estimated results with observed results. By separating what has been reliably measured from what is a completion or an estimate, you make it easier to preserve the reliability of cross-sectional drawings. If necessary, supplementing judgments with on-site photographs or additional records is also effective.

Also, when comparing multiple cross-sections, checking whether any one cross-section has an unusually sparse number of points makes it easier to separate a problem with the extraction settings from a problem with the source data. If surrounding cross-sections have enough points but only one spot is sparse, you should suspect a localized data gap at that location. Conversely, if they are all similarly sparse, you may need to reconsider the extraction width or the filtering/processing conditions.

If you slice without checking point cloud density or gaps, you're likely to end up in a state that won't improve no matter how many times you extract. When a cross-section doesn't appear, it's important to calmly check not only how you slice but whether there are enough points there in the first place.

Cause 6: Extraction only, with no prerequisite for creating cross-sectional drawings

The final major reason people feel they cannot extract cross-section drawings after loading LAS files is that they treat cross-section extraction and cross-section drafting as the same thing. Even if you obtain a cross-section-like display from a point cloud, that does not necessarily become a cross-section drawing that can be used directly in practice. Extraction is simply the stage of retrieving the raw material for drafting; from there you need a process to organize and refine it into a proper drawing. Without this premise, you end up in a situation where cross-sections appear but you “cannot produce a usable cross-section drawing.”

Point-cloud cross-sections always have thickness and variability. Because they display the observation points themselves, they almost never form an ideal single-line section. Both the ground surface and walls exhibit a certain degree of scatter as points. However, the cross-sections required in practice are diagrams organized so that meaningful contour lines, ground-surface lines, break points, and height relationships are easy to read. If you assume that extraction alone completes the task without understanding this difference, you are likely to feel that you cannot produce proper cross-sections.

Especially in terrain cross-sections, it is necessary to consider which line to adopt as the representative line from the thickness of the ground surface in the point cloud. In structural cross-sections as well, you must determine which surface to treat as the true face from the scatter of points. This step is not merely a clean-up but an exercise in clarifying the meaning of the cross-section. Even if the extraction conditions are good, without a policy on what to retain in the drawing, the final deliverable will be unstable.

As a countermeasure, it is important to treat cross-section extraction and diagramming as separate steps from the outset. In the extraction stage, prioritize first whether the points needed as material for the cross section are visible. Then, when transferring into CAD, organize the ground surface line, outline lines, auxiliary lines, and temporary verification lines, and arrange them into a form readable as a drawing. With this premise, you can avoid demanding a perfect single line at the extraction stage and can set conditions more calmly.

It is also effective to decide diagramming rules in advance. If you have rules such as taking lines at the same level of granularity for every cross-section, treating interpolated lines separately, and organizing them so that the elevation reference is clear, it will be easier to achieve consistency when multiple cross-sections are placed side by side. If you decide how to represent them on the spot each time you extract them, the differences between cross-sections will become large and they will be difficult to compare.

In practice, a common situation is that people assume the extraction has failed because the appearance of the extracted points is inconsistent. However, in many cases the extraction actually succeeded, and the only issue was that there was no plan to proceed from there to visualization. When a cross-sectional drawing cannot be produced, you should check not only the extraction conditions but also whether you have a concept for how to convert the extracted data into a drawing.

To stabilize cross-section extraction, you need not only the technique for cutting the section but also the perspective to make it valid as a drawing after cutting. Rather than stopping at extraction alone, treating diagramming as a single continuous process is the final important countermeasure.

Practical Approach to Stabilizing LAS Cross-Section Extraction

So far we've looked at six causes and countermeasures, but what truly works in practice is not memorizing them individually, but proceeding with a stable flow from the start. To ensure LAS cross-section extraction can be carried out the same way every time without hesitation, it's important to standardize the sequence of tasks.

First, clarify the purpose of the cross-section drawing and then decide on the approach to the reference line. If it is clear which cross-section you want to view, it becomes easier to determine the orientation and position of the section line. Next, after loading the LAS, do not extract immediately; check the coordinates, elevation, point cloud density, classification status, and locations of missing data. By getting a feel for the point cloud’s characteristics here, later adjustments to extraction conditions will be considerably shorter.

After that, perform cleanup to reduce unnecessary points in the cross-sections. Adjust the cleanup strategy depending on whether the cross-section is of the ground surface or of a structure, and prioritize removing points that interfere with the cross-section. Rather than immediately extracting all cross-sections, test several provisional cross-sections at representative locations and, while reviewing the line position, orientation, and width, finalize the optimal conditions to minimize rework.

When the conditions are finalized, proceed with the main extraction and review the extraction results in plan, side, and three-dimensional views to confirm that the positional relationships and shapes meet the intended purpose. Based on that, treat section extraction and section drafting as separate processes and organize them into a form usable in CAD. Finally, if you record which positions were cut, at what widths, and how, it will be easier to add sections or re-extract later under the same conditions.

If you follow this workflow, extracting cross-sections becomes considerably more stable. The important thing is not to try to do loading, extraction, and visualization all at once. Proceeding by confirming the conditions one by one at each step will ultimately be faster and lead to more accurate cross-sections. In practical LAS work, it is more important to resolve uncertainties in the earlier steps than to prioritize speed of operation.

Summary

When cross-section extraction fails after loading LAS, the cause is often less the extraction function itself and more assumptions about coordinates and elevations, the position and orientation of the cross-section line, the extraction width, the cleaning up of unnecessary points, overlooking point cloud density or missing data, and insufficient assumptions for generating the cross-section. Simply reviewing these six items in order makes it much easier to isolate what is causing the cross-section to be unstable.

The important thing is not to assume you can get cross-sections right away after loading an LAS. Cross-section extraction is not just slicing the point cloud; it is the process of deciding what you want to see, setting the conditions, removing unwanted objects, extracting at an appropriate width, and finally preparing the result in a form usable as a drawing. If you understand and follow this workflow, you can greatly reduce failures in cross-section extraction.

Furthermore, if you want to make cross-section extraction more stable, it is effective to review not only in-office processing but also which locations on site are recorded and with what intent. If there are photos and records that make it easy to trace cross-section positions and supplementary inspection points afterward, work can proceed much more easily even in situations where decisions are difficult based solely on LAS. As a way to make it easier to keep such geolocated site records, the idea of utilizing iPhone-mounted, high-precision GNSS positioning devices such as LRTK is a good fit. If you want to make the workflow from LAS import to cross-section extraction more reliable and reproducible, it is important to standardize how positions are recorded on site together with establishing procedures in CAD.