1. Introduction
1.1. Research Aim
The Tower of Lopera is a neglected heritage asset situated in a rural context. Despite its current state of abandonment, it constitutes a 14th-century defensive fortification of considerable architectural and historical interest. Beyond its individual value, the tower is part of a broader network of contemporaneous fortifications erected with a common purpose: the defense and repopulation of a volatile frontier zone. Nevertheless, knowledge of this structure remains almost entirely absent—both among the authorities responsible for its protection and among the local population. This research therefore sets out with a dual aim:
Firstly, to generate foundational knowledge about the tower itself, which is essential for any future conservation or intervention measures. In line with the principles established by the 1964 Venice Charter [1], “in all conservation, restoration or excavation work, there must always be precise documentation in the form of analytical and critical reports, illustrated with drawings and photographs.” To this end, contemporary methods of architectural measurement and graphic documentation, such as photogrammetry and laser scanning, are employed—techniques that have become widely adopted in recent years [2,3,4].
Secondly, given that this tower is part of a wider network of similarly neglected fortifications, we aim to propose a replicable working methodology that could be applied to the study of comparable structures. This approach could serve as the basis for coordinated regional research initiatives, and might even prove applicable to other contexts with analogous heritage challenges.
1.2. Historical Background
The conquest of Al-Andalus—the territory of the Iberian Peninsula occupied by Muslim powers from 711 AD—by the Christian peninsular kingdoms was a prolonged and complex process, characterized by shifting alliances and conflicts that did not always follow religious lines. This article focuses on the final phase of that process. In the first half of the 13th century, the Castilian conquest of the Guadalquivir Valley, alongside the Aragonese expansion in the eastern Iberian Peninsula (Levante), effectively reduced the once-mighty Al-Andalus to the Nasrid Kingdom of Granada. This small Islamic state, nestled between the Baetic mountain ranges—home to some of the country’s highest peaks—and the Mediterranean Sea, endured for over two and a half centuries until its final annexation by the Catholic Monarchs in 1492.
A militarized border zone, known as La Banda Morisca, was thus established. This frontier extended from Gibraltar to the Sierra de Cazorla, taking advantage of the natural defenses provided by the region’s mountainous terrain [5]. Today, this historic frontier corresponds to the southern part of the province of Seville and neighboring areas of Cádiz and Málaga (Figure 1a). Chronologically, the period under study falls within the Late Middle Ages, from the mid-13th to the late 15th century.
The expression Banda Morisca is believed to have originated in Seville during the 14th century, much like the term Banda Gallega, which referred to territories under the jurisdiction of Seville that lay across the Guadalquivir River towards Portugal. More precisely, the Banda Morisca denoted the group of frontier lands governed by the Seville council, including, at various times in the 13th century, towns such as Morón, Osuna, and Cote, and, during the 14th and 15th centuries, also Matrera and Arcos [6] (p. 13).
Vestiges of that period persist today in local toponymy: names like Jerez de la Frontera, Arcos de la Frontera, Morón de la Frontera, and Aguilar de la Frontera clearly reference their frontier origins [7]. Beyond this linguistic legacy, however, stand the physical remnants—largely forgotten—of a once-vital defensive infrastructure: the fortifications erected to guard this border. Most now lie in ruins, silent witnesses to the passage of time.
In the early 14th century, driven by fears of a renewed North African invasion, the council of Seville—supported by the Castilian crown and regional nobility—embarked on an ambitious defense program. This initiative demanded logistical efforts far beyond the means of the time and led to the construction of a network of at least 40 defensive towers and minor fortresses strategically positioned across the southern frontier (Figure 1b). These structures, often visually connected to one another, aimed to intercept Nasrid incursions into Christian territory. Only a handful remain standing today; the majority have fallen into advanced states of decay or vanished entirely. Documentation is scant—often limited to old sketches, traveler engravings, or vintage postcards. Nonetheless, the analysis of this graphic material remains a valuable tool for recovering historical knowledge of these sites [8].
The Lopera tower was part of the first defensive line within this system. Its primary function was to protect both the surrounding countryside and the city of Seville from potential Moorish raids originating from the Ronda area. These fortified outposts also served as shelters for the sparse rural populations living in these contested regions, as well as for their livestock [9,10].
Today, most of these towers are abandoned and isolated in rural landscapes. It is likely that many have not undergone significant intervention for centuries. In our view, this situation presents an important research opportunity, particularly given the near-total lack of systematic documentation. Nevertheless, this type of study remains challenging due to the extensive fieldwork required and the difficulty of locating historical references, which are often limited to fleeting mentions in medieval sources.
2. Topographical Context and Layout of the Fortification
The fortification is situated atop a relatively steep hill, with a slightly gentler slope towards the north (Figure 2). The preparation of the terrain for construction does not appear to have posed major challenges, as the structure occupies the highest point of the hill, where several natural rock outcrops are present. This afforded the builders the advantage of a solid geological foundation. In some areas—particularly the steep and rocky southern flank—ashlars can be observed filling cracks and crevices in the bedrock. It is also worth noting that in this southern zone, the surviving outer masonry wall is difficult to distinguish from the natural rock, suggesting that part of the enclosure may have been hewn directly from the bedrock itself.
The fortified complex features an elongated, polygonal plan aligned along an east–west axis. It adapts to the topography of the highest and most defensible part of the hill, while also taking advantage of the small plateau that crowns its summit. The perimeter was once enclosed by a curtain wall, or outer shell, of which only partial sections remain. This wall was clearly laid out to follow the contours of the terrain, adapting to the local geomorphology (Figure 3).
The best-preserved sections of the enclosure are precisely those that face the steepest slopes—namely, the southern and western sides. In contrast, the northern and eastern sectors, which are more exposed and accessible, are the most deteriorated. This condition can likely be attributed to systematic plundering and stone removal over the centuries, particularly during the 19th and 20th centuries, when advances in transportation enabled more efficient extraction and relocation of building materials. It is reasonable to assume that the local population took advantage of the most easily accessible areas.
Scattered stones and ashlars can be observed across the entire hillside, along with accumulations of material in certain locations, appearing as though they had been stockpiled for collection. Despite the degradation, the route of this advanced defensive enclosure remains partially legible, albeit with difficulty in some areas. Its hypothetical reconstruction requires a degree of interpretative reasoning and an understanding of defensive construction logic, especially along the more severely damaged sides.
The thickness of the enclosing wall, or “jacket”, generally exceeds 1.60 m, although this dimension varies from section to section. Traces of a barbican or outer parapet are evident on the western façade, which may have served as an additional line of defense protecting the sole access point to the fortress. This entrance is reached by ascending a series of carved stone steps that lead to a narrow opening, measuring 1.25 m in width. The doorway itself—still largely intact (Figure 4)—is surmounted by a false arch constructed using the corbelling technique through the progressive approximation of courses.
The entrance opens onto a small, almost square courtyard, approximately seven meters per side. It comprises the section of wall containing the entrance gate, the main tower, and two perpendicular walls, one of which has survived. The missing wall—of which only the remains of a buttress (machón) are visible—is thought to have housed the gate providing final access to the fortress’s parade ground, which surrounds the tower [11] (p. 267), This area contains one of the most intriguing features of the complex: its floor lies at a lower level than the rest of the parade ground. This appears to result from an adaptation to the natural topography, as the current level coincides with that of the main entrance. Whether this depression is of natural or artificial origin, it would have made a direct assault through the main gate significantly more difficult.
The parade ground itself is no longer visible today, as the northern and eastern stretches of the outer wall that once defined its limits have disappeared. However, it is possible to deduce its original layout: after a narrowing near the entrance—between the keep and the southern wall—the space likely expanded towards the northern and eastern sectors of the enclosure. Each corner of the perimeter appears to have been flanked by a tower. The towers corresponding to the lost walls have not survived, although a pile of stones at one corner suggests the presence of a structural mass. Collantes de Terán posits that these towers were circular in plan [12] (p. 171), possibly solid cubes without upper chambers. Nevertheless, the surviving towers in the western zone are rectangular and solid, leading us to favor the hypothesis of a consistent rectangular layout for all four.
The two western towers do remain, both of which have a rectangular floor plan, though they differ in size and form. The larger and more complex of the two stands adjacent to the entrance gate, measuring approximately 4 × 3.5 m, and projects markedly from the curtain wall (Figure 5). This tower defended both the entrance and the north-west corner of the fortress, where a barbican was once located. A notable feature is the groove on its north face, which may have been intended for a portcullis. The south-western tower is smaller and projects only slightly from the wall, yet it is the tallest, as it is built atop a substantial rocky outcrop (Figure 6). Both towers are of solid construction and do not contain internal chambers.
In short, the walled enclosure served as the first line of defense against less intense assaults. Defensive action would have been conducted from the upper platforms of the towers—which lacked additional defensive devices—as well as from the curtain walls themselves. Should this perimeter be breached, as might occur in the case of a determined attack, the garrison would retreat into the keep, which could function as a last redoubt or, if necessary, as a true defensive stronghold.
The Main Tower
The keep is situated in a central position within the walled enclosure, though slightly offset to the west. It occupies the highest point of the promontory which, combined with its height of approximately 17 m, makes it the most advantageous vantage point from which to observe the surrounding area. The view extends across the entire northern and western arc—encompassing the countryside of Seville and Cádiz—as well as south and east, up to the Sierra de Grazalema mountain range.
The structure has a rectangular floor plan (12.90 m × 11.30 m), built with excellent ashlar masonry, giving it a solid and monolithic appearance (Figure 7). It stands on a masonry podium approximately one meter high. The walls are between 2.30 and 2.40 m thick throughout, except on the east side, where the staircase is embedded, increasing the thickness to 3.40 m. This considerable mass provides strong defensive capacity, as in the late Middle Ages, artillery was still in its early stages, and there was no effective way to bring down a wall of such caliber except through prolonged siege using long-range projectile weapons—an unlikely scenario in such an isolated location.
The tower comprises a ground floor, an upper floor, and a rooftop terrace, which likely lacked battlements but would have had a protective parapet. Access to the ground floor is via a doorway located at ground level in the south wall. Today, its original appearance is difficult to determine, as material theft has particularly affected this entrance, leaving it heavily deteriorated. If we take the towers of Águila and Aguzaderas as points of comparison—where the entrances are better preserved—it is likely that the doorway was framed with ashlar blocks and featured modillions on either side forming a kind of double lintel. A square hole remains in one jamb, which likely housed the mechanism for a large wooden bolt that would have secured the door.
Once inside, one enters directly onto the ground floor, which is almost square in shape (7.14 × 7.05 m). This space is covered by a barrel vault built in well-cut ashlar masonry, slightly smaller in dimension than that of the outer walls. The vault rests on semicircular arches that project from the wall face and terminate in salmeres, which serve as the transitional element between the arches. The central rings of the vault have been lost, creating an opening approximately 3.5 m in diameter through which the upper floor is visible (Figure 8).
The chamber contains only a single opening: an arrow slit located in the western wall. This narrow window, placed more than four meters above floor level and set within an inward splay, was most likely intended for lighting and ventilation rather than for defensive purposes. The doorway that once led to the staircase would have been located on the opposite wall, but that section of wall has largely disappeared, taking with it the door and the initial section of the stairway. However, the niche beneath the staircase—covered by a pointed ashlar arch and occupying the full width of the stairwell at that level—is largely intact.
The staircase, which provides access to both upper floors, is embedded within the eastern wall to the right of the entrance. It measures just over one meter in width and is roofed with finely dressed lintels supported by modillions, similar in style to those found on the doors and windows. The steps are irregular, with risers taller than the depth of the treads. The upper chamber is accessed via a small landing (1.50 × 1.14 m), which leads through a well-preserved linteled doorway. A loophole opens onto the landing from the north wall, providing light. The second flight, which leads to the rooftop, is structurally similar, although a significant portion of the separating wall remains intact, and only the first seven steps are missing. The rooftop entrance lacks any sort of protective housing and was likely closed by a simple wooden trapdoor.
The upper chamber was probably the principal living space, as is typical in this kind of tower. It resembles the lower floor in shape but is considerably brighter, with openings on all four walls. It is also covered by a barrel vault, which in this case is largely preserved. The window in the southern wall is the largest (1.20 × 2.40 m), centrally placed, and though weathered, is still in acceptable condition. Its double-stepped lateral moldings remain visible (Figure 9), similar to those framing the entrance door. A second window opens in the western wall, resembling the one on the northern side.
The window arrangement supports the theory that the south side was considered the least vulnerable to attack due to the steep slope, allowing for a larger, non-defensive opening on that façade [13] (p. 44).
Of particular interest is the window on the eastern wall (Figure 10), perhaps the most notable architectural element of this level. It is set within a niche in the staircase wall—resembling the niche found on the ground floor, also with a pointed arch but slightly smaller. At the back of the niche is a window with a linteled top, which then narrows sharply into an arrow slit, like its northern counterpart. Above this niche is a small rectangular opening connecting the stairwell to the main chamber. This aperture appears to serve only for visual communication, as it does not open onto the opposite wall that encloses the stairwell.
All doors and windows, both internal and external, appear to have been secured with two-leaf closures, as evidenced by the clearly visible grooves cut into the arch voussoirs—a method typical of other towers of the period.
Finally, the staircase leads up to the roof, now severely damaged by both the elements and human intervention. It likely offered the most accessible source of building materials, contributing to its present state of ruin. The parapet has completely disappeared. Moreover, the proliferation of wild vegetation and nesting birds has further accelerated the roof’s deterioration.
3. Materials and Methods
3.1. Laser Scanner Measurements
A detailed architectural survey is essential for understanding the history and construction of a heritage building—particularly when historical documentation is scarce. The specific aims of each study also dictate the appropriate level of precision and detail required during its execution.
Currently, a wide range of techniques is available for surveying heritage architecture, each with its own advantages and limitations depending on the typology of the elements to be recorded and the nature of the results sought [14].
In the present case study, the main tower is subject to two significant constraints that influence the choice of surveying method. First, its semi-abandoned state and considerable age complicate interior access, which, although possible, must be undertaken with caution. Second, and more critically, its location on uneven terrain with complex orography restricts both physical access to the structure and a comprehensive external view—making it particularly difficult to obtain generalized data.
The most common approaches to dimensional data acquisition today range from discrete data capture using traditional measuring tools or laser distance meters to advanced techniques that enable the generation of comprehensive datasets through point clouds, which are then processed in post-production.
The latter category includes a number of techniques developed to generate three-dimensional point clouds of architectural elements. Traditional terrestrial laser scanners (TLS) have been widely used for this purpose, but they are increasingly complemented—or in some cases replaced—by photogrammetric methods [15,16].
For our study, we considered two principal types of laser scanning technologies, each with distinct capabilities: fixed terrestrial laser scanners (TLS) and mobile mapping systems (MMS). Owing to the specific conditions of our site, we opted for MMS scanning, as it proved better suited to the constraints posed by the site’s accessibility and geometry.
TLS is known for delivering the highest degree of precision, thanks to its highly specialized instrumentation and the stringent planning it requires [17,18]. TLS devices operate from fixed stations strategically positioned around the survey area. These stations capture data radially, with point density decreasing with distance from the scanner. While multiple stations can be combined to improve coverage, this technique becomes impractical for structures with highly irregular geometries or limited access. Comprehensive documentation in such contexts would require a large number of scan positions to achieve uniform coverage.
To overcome these limitations, mobile scanning systems have emerged. These integrate the benefits of laser scanning with the flexibility of motion-based surveying. MMS devices use LiDAR sensors—similar in function to those employed in TLS—that record the distance to surfaces while the operator moves through or around the site.
The key advantage of MMS lies in its movement and positioning algorithms. Using SLAM (Simultaneous Localization and Mapping) technology, MMS systems continuously map their position in space while simultaneously building a point cloud that expands with every step taken [19]. This capability ensures that the scanner maintains a relatively consistent distance and line of sight with respect to the object, yielding more homogeneous and complete point clouds, and reducing the occurrence of blind spots.
However, this same dynamic tracking introduces a potential source of error. SLAM algorithms, whether based on visual cues or inertial measurement units (IMUs), reconstruct the scanner’s trajectory in real time. Over longer scans, this process can accumulate small errors—known as drift—which introduce distortions into the final point cloud. To minimize such inaccuracies, specific scanning protocols are recommended, such as performing closed-loop or circular trajectories, which help mitigate drift and improve overall accuracy.
These capabilities, despite certain limitations, led us to conclude that a mobile mapping scanner was the most suitable option for obtaining sufficiently accurate data from the tower. Its mobility allowed us to navigate around the structure and access difficult-to-reach areas both inside and outside. This flexibility was essential to our case study, as the rugged terrain surrounding the tower rendered the use of fixed scanners, despite their potentially greater accuracy, impracticable.
3.2. Photogrammetry
Digital photogrammetry [20], or structure from motion (SfM), is a technique that enables the generation of point clouds from a series of photographic images. These images must meet a number of requirements and be interrelated in such a way that algorithms can determine the position of hundreds of thousands of points via triangulation [21]. In this manner, photogrammetry can produce 3D reconstructions comparable to those generated by laser scanning, using considerably less expensive equipment. Although the resulting point clouds generally have a slightly lower density than those obtained by laser scanners, in many cases, they are sufficient for the intended purposes.
The images required for photogrammetric procedures are subject to several constraints. First, it is essential to capture a sufficient number of images from varying viewpoints in order to resolve the position equations. These images must also overlap by a certain percentage to allow for the reconstruction of the relative position of all cameras involved.
Furthermore, the pixels within the images must be sufficiently consistent so that the algorithm can identify identical areas across different frames. Homogeneous lighting and careful control of shadows and reflections are therefore crucial. Nonetheless, the success of the photogrammetric reconstruction depends heavily on the context and the equipment used. Exterior photography—especially in bright conditions and on surfaces with pronounced textures—typically offers the most favorable environment for this technique. In such cases, the primary concern is the presence of harsh shadows, which can result in overexposed or underexposed areas with a loss of detail.
Conversely, interior photography presents far greater challenges. Lower light levels necessitate increased sensor sensitivity, which may introduce digital noise that renders the images unsuitable. Since the algorithm relies on detecting similar pixels across images, this distortion caused by noise can result in reconstruction artefacts or even cause the entire process to fail. These considerations directly influenced the application of this technique in our case study. The exterior survey was feasible thanks to the use of UAVs, which, in broad daylight, provided sufficient lighting to capture images from a moving platform without distortion. Special care was taken to schedule the flights when the sun was high enough in the sky to avoid backlighting, which could otherwise cause underexposed or overexposed areas with poor definition.
As for the interior, image capture was possible due to the large opening left by the collapse of the dome, which allowed sufficient natural light to enter. To minimize digital noise and motion blur, photographs were taken using tripods, enabling low ISO sensitivity settings that resolved most of these issues. However, the use of tripods introduced further complications, particularly on uneven floors and during the transport of equipment to the top of the tower. This was especially challenging given the partially collapsed staircase, which made access both difficult and hazardous.
3.3. Case Study
Under the conditions described above, a mobile laser scanner of the GeoSLAM ZEB-REVO type was employed for the various interior surveys. This scanner is capable of capturing 43,200 measurement points per second, with an instantaneous error margin of ±15 mm. These error values refer to the dimensional accuracy at each individual point. However, as noted previously, general acquisition errors—resulting from the scanning environment and the paths taken—can range from 3 to 30 cm across the complete dataset (Figure 11).
As already mentioned, the tower is located at a considerable elevation in a position that is difficult to access, and it is only possible to observe it externally either from close proximity or from a much greater distance. This situation necessitated the adoption of alternative technical solutions for the exterior survey, specifically the use of unmanned aerial vehicles (UAVs), which could capture images from all external viewpoints for subsequent processing.
For the exterior survey, we primarily relied on photographic imagery, supplemented by a limited number of ground-based measurements obtained with a total station. This led to the application of photogrammetric techniques to complement the data acquired from the interior scans.
In the present case study, photogrammetry not only enabled the reconstruction of the element’s exterior but also provided color data for the point cloud (Figure 12). Furthermore, this technique allowed the generation of orthophotographs and other derivative products that were instrumental in producing the final survey. Although not a primary objective, this method also facilitated the reconstruction of the surrounding area, thereby enabling a broader contextual analysis of the tower’s location at the summit of the mound.
It should be noted, however, that successful reconstruction using this technique requires a certain level of expertise, as the quality and accuracy of the results can vary significantly depending on the cameras used and prevailing lighting conditions [22]. In our case, the images were captured using a Mavic Air drone and its integrated camera. Special attention was paid to the fact that the camera employs a so-called “fisheye” lens, which introduces image distortion. These distortions had to be calibrated and corrected during the processing phase in order to achieve accurate results.
Similarly, the position and scale of the point cloud were approximately correct due to the UAV’s GPS capabilities, but they were precisely adjusted using data collected by a total station, which enabled the accurate measurement of various dimensions and reference points on the exterior of the tower (Figure 13). This allowed the structure to be accurately positioned and scaled within the overall model. Additionally, the total station facilitated the measurement of both interior and exterior points, making it possible to align and integrate the external and internal surveys.
Lastly, photogrammetry was also applied to the interior of the tower in addition to laser scanning, as the point cloud generated by the scanner lacked color information. This step was therefore necessary to provide enhanced visual data of the interior wall surfaces (Figure 14). The resulting photogrammetric model was aligned with the laser scan using shared reference points recorded by the total station, enabling the projection of textures onto the laser-derived point cloud. This comprehensive integration also allowed for a comparative analysis of the outcomes generated by each method (Figure 15 and Figure 16).
The interior photographs were captured using a Nikon D5300 fitted with an 18–55 mm Nikkor standard lens. Due to the good lighting conditions inside the tower, no additional equipment, such as flash units, was required. However, a tripod was used to stabilize the camera, thereby minimizing image noise and motion blur.
The precision of both the laser scanning and photogrammetric methods is employed. Specifically, the portable laser scanner (ZEB Revo RT) provides a relative accuracy of ±1–3 cm in field conditions, which was validated through comparison with total station measurements. The photogrammetric survey, processed using Agisoft Metashape, achieved sub-centimeter precision, with a reprojection error below 0.3 pixels and a dense point cloud spacing of 1–2 mm.
The integration of 3D laser scanning and photogrammetric data was achieved through point cloud alignment and visual texture mapping. First, both datasets were imported into CloudCompare, where manual and ICP-based registration was carried out. Secondly, high-resolution orthophotos and textured 3D meshes derived from photogrammetry were overlaid onto the laser scan geometry to enhance visual readability.
4. Data Cleaning and Discretization
The application of these techniques provided highly valuable results; however, further processing was required to transform them into structured and usable information. The point clouds obtained do not consist of discrete hierarchical elements, but rather form dense, uncategorized aggregations of spatial data. While visually informative, these datasets do not directly yield the planimetry necessary for architectural documentation, such as floor plans and sections. As a result, manual extraction of such data becomes essential in order to produce precise architectural representations.
Although various attempts have been made to automatically extract basic geometrical elements—such as cubes, spheres, and cylinders—from point clouds, current methods still fall short of the precision and complexity required for structures such as the one in this study. The intricate geometry of the tower cannot be accurately represented by simple forms, making manual intervention indispensable.
Nevertheless, the point clouds obtained through laser scanning and photogrammetry (Figure 17) were refined through two primary processes:
4.1. Alignment of the Point Cloud Dataset
One of the most critical and complex tasks involved the spatial alignment of the partial datasets. These comprised the exterior survey and the two internal levels of the structure. While the interior levels were separately captured using the mobile laser scanner, photogrammetric reconstruction managed to produce a unified model, thanks to the large central gap in the vault, which allowed visual continuity between the lower and upper areas. This enabled the algorithm to generate a single, coherent volumetric reconstruction.
The alignment process was guided by control points obtained with the total station. These included coordinates from both the external façades and several internal locations, notably the station’s own placement within the interior.
A number of challenges were encountered in acquiring common reference points. The narrowness of the entrance, the steep slope immediately surrounding the tower, and the deteriorated state of the stairs to the upper level posed significant obstacles to the accurate positioning of the total station, making data acquisition a laborious task at times.
Ultimately, once the respective models were loaded and the common reference points were introduced into each dataset, we were able to achieve a coherent alignment of the complete survey. This allowed the generation of accurate sectional views, facilitating subsequent analyses such as wall thickness measurement and interior–exterior spatial relationships.
4.2. Obtaining CAD Planimetry and Flat Images
The planimetry of the final model was developed using AutoCAD 2024 software. To achieve this, the aligned point clouds were imported into 3D modeling software, from which the main outlines of the structure were traced. A selection was made of the relevant points required to generate a balanced representation that combined both detail and the regularization of planar areas or smooth curves, enabling the geometry of the heritage element to be reconstructed [23].
Although the process was predominantly manual, planar sections of the point clouds or of the meshes generated from them were obtained for some of the views. These meshes, formed through triangulation of the original points, are produced by algorithms whose parameters must be carefully chosen. It is essential to find a balance that corrects minor errors inherent to the technique, while avoiding excessive loss of detail and, therefore, of valuable information.
Once the models were translated into planar and discretized geometry, a series of geometric analyses typical of this type of study were applied. These analyses aimed to identify the original layout patterns used in the construction of vaults, staircases, and other architectural features.
In addition, for the analysis of masonry surfaces, this type of model allows the extraction of orthogonal, textured images. These images are created by projecting the textures from the photographs onto the previously generated meshes, thereby enabling the production of flat elevations of the walls. These images capture a high level of detail, which proved fundamental in our case study, enabling an in-depth reading of the construction phases visible on the wall surfaces [24,25].
Finally, by combining these outputs, we generated planimetric drawings—primarily sections with the projection of rear-facing walls. The resulting documents offer high informational value, allowing the simultaneous appreciation of many different aspects of the heritage asset (Figure 18 and Figure 19).
5. Conclusions
5.1. Conclusions on the Methodology Used
In this case study, which focuses on one of the most complex fortifications of the Banda Morisca, we conclude that the most effective approach to the digital graphic analysis of such a structure is the combination of close-range photogrammetry with interior imagery and detail shots obtained using a DSLR camera. These two sets of images—processed separately—can be unified by referencing the points captured with the total station, thereby generating a single, cohesive point cloud of the entire structure. The geometric accuracy achieved is comparable to that obtained through mobile 3D scanning, which is particularly advantageous in contexts where the use of terrestrial laser scanning (TLS) is hindered by terrain constraints. Furthermore, this approach has the added benefit of providing textured outputs.
An additional strength of this methodology is the capacity to survey all elevations of the tower, enabling the reconstruction of its historical stratigraphy and facilitating the development of preventive conservation strategies (Figure 20).
Moreover, producing a high-definition point cloud of this nature proves especially valuable for research projects involving heritage assets dispersed across rural areas—locations where drone operation is generally permitted and access is often challenging. This enables the creation of accurate digital models of each asset, allowing for more detailed analysis and typological comparison, as well as more precise geometric studies in subsequent phases.
The objective of this work is to optimize the methodological approach within the constraints of our limited resources. As there is currently no dedicated budget for this heritage project, and even if funding were secured, the human and financial resources would likely remain modest given the number of sites requiring survey. Thus, our intention is to present a realistic and applicable methodology based on the tools and means available.
5.2. Conclusions on Dimensions
Although dimensional analysis was not the primary objective of this study, several noteworthy observations have emerged. These are not fully developed here, as they fall somewhat outside the main focus of this paper and would require further elaboration beyond its scope. We intend to explore them in greater depth in a subsequent publication.
One of the most intriguing aspects of this fortification is the vaulting of the two floors of the tower. Both vaults are still present in the Lopera tower, albeit in a semi-ruined condition. They are constructed of ashlar and span square rooms measuring over seven meters per side—an unusual feature in buildings dating from the 13th–14th centuries. The three-dimensional photogrammetric analysis has allowed us to investigate previously posed questions, such as whether the vaults are truly hemispherical or segmental, whether their centers share a common axis, or whether they are displaced.
Our analysis confirms that the ground floor vault is hemispherical, while the upper floor vault is segmental and displaced from the central axis. This is primarily due to the reduced height of the upper perimeter wall, which allows the inscription of a shallower spherical form.
When comparing these findings with plans generated using conventional techniques, we can clearly observe their limitations relative to those derived from the point cloud data (Figure 19). The latter provide far greater precision and offer a sound basis for evaluating the applicability of this digital measurement methodology to other similar structures.
It is no coincidence that the tower’s measurements appear arbitrary when expressed in metric units. The decimal metric system was not yet in use at the end of the Middle Ages, and a range of traditional measurement units was employed across the Iberian Peninsula. In late medieval Andalusia, the so-called vara burgalesa, later referred to more broadly as the vara castellana, seems to have been widely used. This unit, which was officially standardized across Castile in the 16th century under Philip II, measured approximately 0.8359 m. It was subdivided into: paso (5/6 of a vara), 1/2 vara, pie (1/3 of a vara), and palmo (1/4 of a vara).
Evidence from 13th–14th-century religious architecture in Córdoba, notably the Fernandine churches constructed by builders from Burgos [26], indicates widespread use of the vara burgalesa, along with its multiples and submultiples. Given the geographical and chronological proximity, it is plausible that the same system was applied in Seville. Accordingly, we converted our metric measurements to this historical unit in order to explore potential proportional relationships inherent in the original construction.
For this purpose, we prioritized interior dimensions and wall thicknesses, as these were more likely to have been deliberately designed in whole units. External dimensions, on the other hand, were more likely to reflect constructional contingencies and thus exhibit less regularity.
The interior room of the Lopera tower measures six metric units, which can be interpreted as six vara burgalesa mayorada. The vara mayorada—used to measure larger spans—corresponded to the diagonal of a square with one-vara sides (i.e., vara × √2). The use of this irrational root was a known medieval geometric strategy, either applied directly in design or through simplified ratios such as Lechler’s 7:5 proportion [27].
The wall thickness measures two varas, reinforcing the hypothesis that both interior dimensions and structural thicknesses were conceived using the same unit system.
Conceptualization, J.F.M.R. and I.A.; methodology, J.F.M.R.; software, J.R.M.; validation, J.F.M.R., I.A. and J.R.M.; formal analysis J.F.M.R.; investigation, J.F.M.R.; resources, I.A.; data curation, J.R.M.; writing—original draft preparation, J.F.M.R.; writing—review and editing, J.F.M.R.; visualization, J.F.M.R.; supervision, I.A.; project administration, I.A.; funding acquisition, J.F.M.R. All authors have read and agreed to the published version of the manuscript.
The original contributions presented in this study are included in the article. Further inquiries can be directed to the corresponding author.
The authors wish to express their sincere gratitude to the Serratosa family for granting access to their rural estate, which allowed us to enter the fortification and conduct the fieldwork essential to this study. This research has been carried out within the framework of the HUM-799 research group, Strategies of Heritage Knowledge, and the Department of Architectural Graphic Expression (EGA), to which two of the authors are affiliated.
The authors declare no conflict of interest.
Footnotes
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.
Figure 1 (a) Location map of the so-called “Banda Morisca”. (b). The network of castles of the “Banda Morisca”.
Figure 2 Fortress situation on the ground.
Figure 3 Ground floor of the fortress of Lopera. Made by the author.
Figure 4 Access to the castle is located on the western elevation of the outer enclosure wall.
Figure 5 Wall closing the access to the castle, located in the northern part of the castle.
Figure 6 The south-west tower.
Figure 7 External image of the Lopera Tower.
Figure 8 Interior image from ground floor.
Figure 9 Window inside a niche, located on the upper floor just above the staircase. The entrance and exit of the room can be seen on both sides.
Figure 10 Upper window on the south façade, just above the entrance door. The side modillions can be seen, as well as the absence of the matacan.
Figure 11 Resulting alignment between the interior and exterior point cloud datasets.
Figure 12 Virtual reconstruction of the Lopera Tower using drone-based photogrammetry.
Figure 13 Interior measurements being taken using the total station. A section of the wall is visible through the doorway opening.
Figure 14 Image of the tower’s interior derived from the point cloud produced using interior photographic data.
Figure 15 (left). Route undertaken with the mobile scanner to generate the point cloud by means of the SLAM algorithm. (right). Photograph positions used for point cloud generation through photogrammetry.
Figure 16 Sectioned point cloud taken with MLS and interior shot by photogrammetry.
Figure 17 (Left): Textured exterior point cloud obtained from drone imagery. (Center): Photogrammetric model of the interior side. (Right): Point cloud generated through mobile laser scanning.
Figure 18 The two floors of the keep, with own elaboration from CAD software.
Figure 19 Section obtained from the final point cloud, produced using CAD software.
Figure 20 Elevations of the main tower, produced by CAD software. (1) East Elevation. (2) North Elevation. (3) West Elevation. (4) South Elevation.
1. The International Charter for the Conservation and Restoration of Monuments and Sites, Known as the 1964 Venice Charter, Adopted by ICOMOS in 1965; Article 16 International Council on Monuments and Sites: Paris, France, 1965.
2. Li, Y.; Li, Z.; Song, Y. The particularity of aerial photogrammetry for architectural heritages by UAV. Proceedings of the 2012 2nd International Conference on Remote Sensing, Environment and Transportation Engineering, RSETE 2012; Nanjing, China, 1–3 June 2012.
3. Castilla, F.J.; Ramón, A.; Adán, A.; Trenado, A.; Fuentes, D. 3D Sensor-Fusion for the Documentation of Rural Heritage Buildings. Remote Sens.; 2021; 13, 1337. [DOI: https://dx.doi.org/10.3390/rs13071337]
4. Elsaid, M.E.; Ayoub, M.; Hassan, H. Scan-to-Building Information Modelling vs. HBIM in Parametric Heritage Building Documentation. IOP Conference Series: Earth and Environmental Science, Proceedings of the Simulation for Sustainable Built Environment, New Cairo, Egypt, 28–30 November 2019; Institute of Physics Publishing: Bristol, UK, 2019; Volume 397, 012015.
5. García Fernández, M. El Reino de Sevilla en Tiempos de Alfonso XI: (1312–1350); Diputación Provincial: Sevilla, Spain, 1989.
6. García Fernández, M.; Mata Marchena, J. La banda morisca durante los siglos XIII, XIV y XV. Actas de las II Jornadas de Temas Moronenses; Fundación Municipal de Cultura Fernando Villalón: Morón de la Frontera, Spain, 1996; 13.
7. Martín Gutiérrez, E. Análisis de la toponomía y aplicación al estudio del poblamiento. El Alfoz de Jerez de la Frontera Durante la Baja Edad Media. Historia, Instituciones, Documentos; Editorial Universidad de Sevilla: Seville, Spain, 2003; pp. 257-300.
8. Molina Rozalem, J.F. Arquitectura Defensiva en las Fronteras del Reino de Sevilla Durante la Baja Edad Media; Subdirección General de Publicaciones y Patrimonio Cultural. Ministerio de Defensa: Madrid, Spain, 2016.
9. González Jiménez, M. La Repoblación de la Zona de Sevilla Durante el Siglo XIV; Secretariado de Publicaciones, Universidad de Sevilla: Sevilla, Spain, 2001; Volume 60.
10. Sánchez Saus, R. Poder urbano, política familiar y guerra fronteriza: La parentela de Alonso Fernández Melgarejo, veinticuatro de Sevilla y alcalde de Zahara. Andalucía Entre Oriente y Occidente (1236–1492). Actas del V Coloquio Internacional de Historia Medieval de Andalucía; Diputación de Córdoba: Córdoba, Spain, 1988.
11. Rojas Gabriel, M. La torre de Lopera, arquetipo de la Banda Morisca. Rev. Arqueol. Mediev. Española; 1987; 2, 267.
12. Collantes de Terán Delorme, F. Los Castillos del Reino de Sevilla; Diputación Provincial Diputación Provincial: Sevilla, Spain, 1952; 171.
13. Molina Rozalem, J.F. Fortificaciones de la Banda Morisca en la Provincia de Sevilla; de Sevilla, D. Servicio de archivos y publicaciones: Sevilla, Spain, 2017.
14. Gil-Piqueras, T.; Rodríguez-Navarro, P.; Cabrera Revuelta, E.; Gandía Álvarez, E. Preliminary Studies of the Coastal Defenses of Cullera’s Cape Built during the Spanish Civil War: From Historical Study to Formal Analysis. Heritage; 2022; 5, pp. 3032-3048. [DOI: https://dx.doi.org/10.3390/heritage5040157]
15. Cavalagli, N.; Gioffrè, M.; Grassi, S.; Gusella, V.; Pepi, C.; Volpi, G.M. On the accuracy of UAV photogrammetric survey for the evaluation of historic masonry structural damages. Procedia Struct. Integr.; 2020; 29, pp. 165-174. [DOI: https://dx.doi.org/10.1016/j.prostr.2020.11.153]
16. Angulo Fornos, R. La fotogrametría digital: Una herramienta para la recuperación de arquitecturas perdidas. Torre del Homenaje del Castillo de Constantina. Rev. Digit. Virtual Archaeol. Rev.; 2013; 4, pp. 140-144. [DOI: https://dx.doi.org/10.4995/var.2013.4355]
17. Grussenmeyer, P.; Landes, T.; Voegtle, T.; Ringle, K. Comparison methods of terrestrial laser scanning, photogrammetry andtacheometry data for recording of cultural heritage buildings. Int. Arch. Photogramm. Remote Sens. Spat. Inf. Sci.; 2008; 37, pp. 213-218.
18. Jo, Y.; Hong, S. Three-Dimensional Digital Documentation of Cultural Heritage Site Based on the Convergence of Terrestrial Laser Scanning and Unmanned Aerial Vehicle Photogrammetry. ISPRS Int. J. Geo-Inf.; 2019; 8, 53. [DOI: https://dx.doi.org/10.3390/ijgi8020053]
19. Benavides López, J.A.; Vera, J.A.B. Evaluación de los Sistemas de Mapeo Móvil (MMS) en la documentación gráfica del tholo de El Romeral (Conjunto Arqueológico Dólmenes de Antequera). Arqueol. Arquit.; 2020; 17, e095. [DOI: https://dx.doi.org/10.3989/arq.arqt.2020.004]
20. FARO Technologies. Available online: https://www.faro.com/en/Resource-Library/Article/What-is-SLAM (accessed on 20 September 2024).
21. Rodríguez-Navarro, P.; Wragg, E.; Verdiani, G.; Gil-Piqueras, T. Modelling the last of the “Movies”: Discussion and digital survey of the Eothen formerly ML286. Virtual Archaeol. Rev.; 2021; 12, pp. 57-72. [DOI: https://dx.doi.org/10.4995/var.2021.14543]
22. Sun, Z.; Zhang, Y. Using Drones and 3D Modeling to Survey Tibetan Architectural Heritage: A Case Study with the Multi-Door Stupa. Sustainability; 2018; 10, 2259. [DOI: https://dx.doi.org/10.3390/su10072259]
23. Martín Talaverano, R. Graphic documentation of historical buildings: Principles, applications and prospects. Rev. Arqueol. Arquit.; 2014; 11. [DOI: https://dx.doi.org/10.3989/arq.arqt.2014.014]
24. Parfenov, V.; Igoshin, S.; Masaylo, D.; Orlov, A.; Kuliashou, D. Use of 3D Laser Scanning and Additive Technologies for Reconstruction of Damaged and Destroyed Cultural Heritage Objects. Quantum Beam Sci.; 2022; 6, 11. [DOI: https://dx.doi.org/10.3390/qubs6010011]
25. Angulo Fornos, R.; Pinto Puerto, F.; Rodriguez Medina, J.; Palomino, A. Digital Anastylosis of the Remains of a Portal by Master Builder Hernán Ruiz: Knowledge Strategies, Methods and Modelling Results. Digit. Appl. Archaeol. Cult. Heritage; 2017; 7, pp. 32-41. [DOI: https://dx.doi.org/10.1016/j.daach.2017.09.003]
26. García Ortega, A. Las parroquias medievales cordobesas: Su traza a la luz de Villard. Expresión Gráfica Arquit. Rev. Expresión Gráfica Arquit.; 2002; 7, pp. 27-35.
27. Ruiz de la Rosa, J.A. De geometría y arquitectura. RA Rev. Arquit.; 2001; 3, pp. 22-32. [DOI: https://dx.doi.org/10.15581/014.3.25970]
© 2025 by the authors. Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License.
Details
; Medina, Jesús Rodríguez; Acosta, Ignacio