Progress in industrial photogrammetry by means of markerless solutions

The main objective of this thesis is to develop and characterize the scope of a novel and edge-cutting approaches based on markerless photogrammetric solutions. Hence, the presented study is focused on going beyond the state of art limitations of industrial photogrammetry avoiding or reducing considerably the employment of artificial targets.

Currently, industrial systems appeal to artificial targets (attached or projected to the part surface) in order to carry out the measurements with high automation and confidence level. Nevertheless, some industrial scenarios are demanding markerless approaches to improve manufacturing efficiency. The main advantages arise both on the avoidance of part or scenario preparation and on the simplification of measuring procedures reducing measuring time consumption and consequently the cost regarding measuring services.

Current limitations and performed research to overcome these technology barriers are presented in Chapter 2. As stated in the review of industrial solutions in Chapter 3, trending manufacturing sectors are demanding more and more flexible and fast measuring solutions that deal with photogrammetric capabilities both for static and dynamic scenarios.

The content of the research is presented and divided as follows. Chapter 1 is a summary of the thesis explaining the different chapters and their content. Chapter 2 introduces and emphasizes the needs of industrial manufacturing processes and how measuring solutions are required to improve current approaches. In fact, a brief description of focused research approaches is mentioned as an alternative amendment. Moreover, it presents the objectives of the research and layout initial hypothesis to address these goals which will the basis for final conclusion and discussion.

Chapter 3 explains and describes the state of art (SOA) of industrial photogrammetric solutions going through existing solutions and approaches in detail. From offline to in-line approaches, an overview and classification of market technologies are presented and required enhancements suggested which the reference points of this thesis are.

As the photogrammetric problem covers multiple concepts and principles to figure out and achieve 3D object information, Chapter 4 deals with the main description of overall concepts and processing steps which are required to reconstruct 3D object information from 2D imagery data. Indeed, the presented workflow is the same for target-based and markerless measuring strategies. This is not expecting to be a detailed description of each term and method, but an overview of problematic and a brief description of main aspects accompanied with accurate references. Besides, meaningful contribution of this thesis is added in some processing steps as a value-added approach.

After presenting the SOA and the main photogrammetric methods applied in common industrial solutions, developed and assessed case studies are introduced in Chapter 5. Three case studies are described and their scope, as well as suitability, has been determined for empowering current industrial applications. For each case study and introduction and main results are introduced as a summary.

Next Chapter 6 deals with the main results and conclusions giving an overall reflection of achieved objectives regarding planned ones. It summarizes the real applicability of developed and studied measuring procedures for real application as well as possibilities of improvement and challenges to be solved in the coming future.

Previously introduced principles and methods are deeply referenced for a deeper lookup and comprehension in Chapter 7. The goal of these bibliographic references is to enable further information search to check the many concepts involved in any photogrammetric solution. Regarding the case studies, each one is independently referenced with appropriate and recent complementary studies.

Finally, Chapter 8 comprises the annex where the cross-reference and citation of published articles is encountered as well as published peer-reviewed full papers. Moreover, interesting web links are provided focusing on photogrammetric fundamentals and measuring capabilities offered by industrial solutions.

A summary at a glance of the employed structure is defined in the following list:
Chapter 1. Summary of thesis and main chapter presentation.
Chapter 2. Overall context and introduction of industrial needs, objectives, and motivation of the thesis and initial hypothesis
Chapter 3. Review of state of art of industrial close-range photogrammetry
Chapter 4. Overview of materials and methods and main contributions
Chapter 5. Description of published case studies and obtained results
Chapter 6. General conclusions and future work
Chapter 7. Bibliographic references
Chapter 8. Annex

The results of each case study are promising and fit to purpose for presented industrial applications, however, further research is still to go in order to replace target-based photogrammetric solutions with markerless ones. Improvements in terms of accuracy enhancement, data handling and image processing strategies, higher computation efficiency and speed in data processing algorithms, standardization of measuring procedures as well as uncertainty assessment studies are still pendant. Simulation of measuring procedures, which is not usually established, is another key point that can be used to foresee measuring process suitability for inspection purposes.

Therefore, as photogrammetric techniques continue developing day by day, edge-cutting data processing procedures and integrated novel and smarter solutions will appear on the coming future covering existing needs and challenges.