In photogrammetry, the quality of the input images is paramount. While using high-quality cameras and lenses always helps, any scan can be enhanced processing the data in image editing software, cleaning up any optical artifacts and improving picture quality. Optical artifacts – imperfections introduced by camera lenses – can compromise the accuracy and and geometry of the resulting models. While a lot of these imperfections are hard to avoid when shooting, they usually just require one click in an editing software to remove, improving the final output while also ensuring a smoother and more efficient workflow. One important note to keep in mind is that this can only be done by shooting in RAW.
Chromatic aberration, also known as color fringing, usually occurs at a border between two contrasting colors, such as the edge of a dark object against a light background. It manifests as red or blue “halos” outlining the high-contrast edge, and arises because lenses bend light of different wavelengths (colors) at slightly different angles. A good introduction into what chromatic aberration is and how to fix it in Adobe Photoshop Lightroom can be found here.
If this optical artifact is left in the input images of a 3D model, it can confuse the software’s feature-matching algorithms, leading to blurred textures and inaccurate geometry. When corrected, these fringing effects are removed, creating sharp, clean edges that improve both the visual and geometric fidelity of the 3D model.
Beyond correcting chromatic aberration, optimizing lighting and contrast in your images is another critical step in pre-processing for photogrammetry. Images with poor lighting or low contrast can obscure fine details, making it difficult for photogrammetry software to identify key visual cues. These visual cues are essential for accurate feature matching and precise geometry generation, allowing the software to create a dense point cloud.
Adjusting highlights, shadows, and mid-tones in editing software helps reveal subtle surface details that might otherwise go unnoticed. This step is particularly valuable when scanning objects with monochromatic or low-contrast surfaces, such as a single-color piece of clothing. By enhancing contrast and brightness, the software can better interpret and pick out unique visual points, resulting in more spacial data to work with. This enhances the textures and features of the object, resulting in a 3D model with improved geometric accuracy and sharper details.
Pre-processing images also plays a crucial role in creating textures that are true to life. Adjustments to lighting and contrast can make a significant difference in how colors and fine details are captured. For instance, black clothing can appear gray or washed out in poorly lit images, while high-quality edits can restore its rich, deep color. Similarly, tiny visual details such as stitching, fabric texture, or brand tags become more pronounced when the contrast and exposure are fine-tuned.
This attention to detail ensures that the resulting textures are not only sharper but also more realistic, capturing the true essence of the scanned object. When applied to 3D models, these textures enhance the overall visual fidelity, making them more engaging and accurate representations of their real-world counterparts.
Pre-processing images for photogrammetry is an easy yet impressively effective way to improve the quality of your 3D models without having to change any hardware. Addressing chromatic aberration and optimizing lighting and contrast ensures that your images provide the software with the best possible input data. This not only improves the precision and geometric fidelity of the resulting models but also enhances the sharpness and realism of textures.
While these steps can significantly enhance your results, it’s important to note that nothing replaces the value of a high-quality photogrammetry rig. By adding these steps to the workflow of our scanners, we are able to deliver ever better 3D models for any application. If you are interested in seeing it for yourself, don’t hesitate to contact our team!