Pattern Analysis Software Tools
The research of ancient written artefacts results in an ever-increasing amount of digital data in different forms, ranging from raw images of artefacts to automatically generated data from advanced acquisition techniques. The manual analysis of this data is typically time consuming and can be subject to human error and bias. Therefore, a set of Pattern Analysis Software Tools (PAST) has been developed for the automatic analysis of visual and tabular patterns in the research data from the study of ancient written artefacts. These software tools have been conceptualised and designed by Hussein Mohammed to facilitate a more efficient study of written artefacts and to help scholars benefit from the rapid advancements in the fields of pattern analysis and artificial intelligence. Furthermore, these tools can provide new insights which can only be derived from the statistical analysis of research data. Each tool in PAST is developed and tested in close collaboration with experts from relevant fields of research in order to ensure its usability and applicability to actual research questions.
Some of these tools are implemented by other members of the VMA lab.
All software tools are licensed under the Creative Commons Attribution Non Commercial No Derivatives 4.0 International, and are intended for research and educational purposes only.
Handwriting Analysis Tool (HAT)
HAT3 is a software tool that can be used to analyse handwriting styles. Multiple and different handwriting styles can be analysed concurrently and sorted according to their similarity to a questioned or unknown style. A similarity score can be calculated for each predefined style to create a relative comparison between them with respect to an unknown style.
Visual-Pattern Detector (VPD)
The VPD is an efficient and easy-to-use software tool for pattern detection. This tool can be used to recognise and allocate visual patterns (such as words, drawings and seals) automatically in digitised manuscripts. The recall-precision balance of detected patterns can be controlled visually, and the detected patterns can be saved as annotations on the original images or as cropped images depending on the needs of users.
ScriptSight
ScriptSight is a software tool that can be used to explore and sort collections of document-page images using their AI-generated ‘computational visual catalogues’. Multiple visual attributes, such as text orientation, used writing implements (pencil versus ink), and text colour, can be filtered concurrently to isolate pages matching the specified criteria. Matching pages can be previewed as thumbnails with optional overlays indicating detected text regions, and and saved as organised, date-stamped folders for review or further processing.
Line Detection Tool (LDT)
The main goal of the LDT is to analyse images of writing supports in order to detect lines (such as sieve imprints or papyri fibres) and estimate their density. These detected lines form a pattern, which can be used as a distinctive feature of the writing support.
Text-Lines Counter (TLC)
The main goal of this software tool is to detect, count and mark the text lines in images of handwritten manuscripts. This version contains features to detect vertical text-lines and bright text-lines on darker background.
X-ray Fluorescence Data Analysis Tool (XRF-DAT)
The main goal of the XRF-DAT is to analyse tabular data generated by X-ray Fluorescence (XRF) spectroscopy in order to ease and speed up the processing and evaluation of data obtained when analysing written artefacts; in particular their inks, pigments and writing supports.
Artefact-Features Analysis Tool (AFAT)
The main goal of AFAT is to calculate statistical information from manually generated tabular data which consists of distinctive features of artefacts. The value of each feature in these tables is represented by a positive integer, which describes the particular variant of this feature in an artefact. Furthermore, each variant can have any number of variations, represented by an alphabetical letter. The combination of variant and variation can be used to describe the observed version of a given feature in an artefact.
Visual Similarity Annotation Tool (V-SAT)
V-SAT is designed to streamline the organisation of image collections. Whether you have a structured folder hierarchy or a jumble of unsorted images, the tool simplifies the process. Once organised, the software enables users to export their curated collections. The main goal of this tool is to enable experts to create training datasets for similarity-based models such as siamese networks and supervised clustering.