Main

Stamp search 1.2

2019-07-25T00:58:21Z

The stamp classification cannot be done with the machine learning technologies as the image recognition is particularly difficult if images are not taken in specific light conditions and because the number of available images for training the neural network was too small. In any case, the classification may be done using the information that can be directly read by the user: the visually recognisable letters on the sherd.

As for the previous classification tools, the first step for stamp classification is the acquisition of the stamp image from a gallery or directly with the camera as shown in the picture below.

Once the image is ready, the user would move to the search interface where he/she may use the search technique based on text patterns to isolate the stamps containing the letters that the user may visually identify in the image.

The search engine allows making searches including wildcard characters in order to match the text (for “Name stamps”) contained in the stamp itself. The scenario is the one in which the archaeologist finds a sherd containing a stamp but only part of the name is readable or almost readable. To search a pattern like “ACCU …. R . S” (dots represents unreadable characters) the search engine allows to represent it as “ACCU%R_S” meaning “ACCU” followed by any character, then “R” followed by one character and finally “S”. The searching for this pattern allows identifying the stamp “ACCU ILARUS”. The wildcard characters that may be used in the keyword field are the following:

% for zero, one, or multiple unknown characters
_ for a single unknown character

To activate the pattern search on the stamp text, the first character of the pattern should be #.

Examples of patterns may be the following:

Typing "#a%" finds any values that start with "a"

Typing "#%a" finds any values that end with "a"

Typing "#%or%" finds any values that have "or" in any position

Typing "#_r%" finds any values that have "r" in the second position

Typing "#a_%_%" finds any values that start with "a" and are at least 3 characters in length

Typing "#a%o" finds any values that start with "a" and ends with "o"

The search is case-insensitive.

Synchronisation 1.1

2019-07-25T00:55:13Z

The synchronisation process is seamlessly carried on in the background each time the device is connected to the internet via WI-FI or 3G/4G. The hidden behaviour of the synchroniser is useful in standard conditions because everything happens without the need of user intervention, but if something goes wrong while using the app the users may be doubtful about the actual status of the collected data.

Tapping on the sync button on the top right corner of the Dashboard (see left image below) the user will access the Synchronisation monitor (see right image below). The page lists the main synchronisation actions (upload on the server, delete from the server, download from the server) on the relevant entities collected by the user (site, sherds, depictions, etc.). When everything is fine, the cloud icon on the left, for each line, will be green. If something goes wrong (for example connection lost during sync) or if the app is currently used in offline mode, the clouds becomes red for each sync action not yet finalised.

Shape recognition 1.1

2019-07-25T01:18:18Z

The shape recognition is based on the profile of a sherd fracture represented in SVG format (D6.3). The process starts with an image of the sherd focused on the fracture and provides tools to extract the scale factor, the external and the internal profile of the potsherds. The extracted data are sent to the API Server that composes the SVG file and sends it to the Shape Recognition Model for a prediction. The results are enriched with database information and sent back to the App.

A call to the recogniser will result in 5 possible ceramic types, each with a specific probability, that the user may evaluate and eventually select and save. The overall process consists of the following sequential steps:

Step 1: Image selection and manipulation
Step 2: Definition of the scale factor
Step 3: Extraction of the external profile
Step 4: Extraction of the internal profile
Step 5: Evaluation of results
Step 6: Manual classification (alternative)
Step 7: Save the info in a sherd (for authenticated users)

The image to be used to sketch the profile may be selected from the device galleries (i.e. previously acquired images) or taken directly with the camera. Once selected, the image may be manipulated using the tools provided by the pen button on the right bottom corner of the screen.

The expected manipulations may be:

Flip horizontal
An important requirement is the in the final sketch the external profile must be on le left-hand side. The horizontal flipping of the image may be used at this purpose.
Rotation
The sherd fracture should be oriented in order to follow the real vessel orientation for improving the prediction probability.

It isn’t necessary to make any other special manipulations of the image because it will not be used as input to the prediction.

The definition of the scale factor is important in order to scale the image to the dimension of real vessels, helping the recognition tool in its search and retrieval task. To define the scale factor the user will be asked to select 2 points in the image and assert the distance in real life. For this reason, users must take pictures of the potsherds always with a ruler, that will have to be placed at the same level as the fractured profile.

Once the scale factor is defined the user may step forward to the extraction of the external profile. The process consists of using a finger to choose points along the external profile of the fracture. The density of the points depends on the complexity of the profile, it is important to capture all the discontinuities and curves of the profile. A smooth outer profile can be sketched with a few points. It is possible to clear the whole sketch or undo points sequentially.

After the extraction of the external profile, the user should provide the internal one using the same technique.

In step 5, the user may evaluate the results of the prediction. Any prediction consists of 5 ceramic types with different probabilities. Is important to note that, as ceramic types are structured in classes and subclasses, the prediction engine tries to identify only the top level classes as discussed in D5.1. For example, the class “Conspectus form 34” (see image above) is the parent of “Conspectus form 34.1” and “Conspectus form 34.2” that differ one from the other of small bits that cannot be easily captured by the neural network but that are also difficult to distinguish by humans.

As already explained for the decoration recognition, if the user isn’t happy with the prediction may proceed manually tapping on the button “Nothing good? Tap to add manually” and enter the visual matching interface. In the case of shape recognition, the comparison is between the sherd image and the drawing associated with the ceramic types in the Reference Database (see picture above).

Image recognition 1.1

2019-07-25T01:10:06Z

The automatic recognition of the decoration class of sherd starts from the image of the sherd that may be taken directly through the camera or from the galleries of the device. The process consists of the following steps:

Step1: Select the image
Step2: Modify the image using the manipulation tools
Step3: Isolate the sherd in the image
Step4: Evaluate the results
Step5: Save the results (if the user is logged in and the results are good)
Step6: Manually classify the image by visual comparison

If the image needs to be adjusted, the user may use the tools available in the menu button (pen on brown background for flipping (horizontally, vertically) rotating and cropping.

If the image needs to be adjusted, the user may use the tools available in the menu button (pen on brown background - see image below) for flipping (horizontally, vertically) rotating and cropping.

Once the image is good, tapping on “Classify” the user moves to the second step: isolating the sherd (see image below). In this step, the goal is to obtain a clean image of the sherd where the background does not influence the recognition model. Usually, simply selecting the area containing the sherd will be enough but some time it may be necessary to remove the background with more advanced tools. The Step 2 interface provides an automatic background removal tool (the “Process” button), the identification rectangle (the most used) or a technique based on telling the GrabCut (see D6.3) algorithm which areas should remain (sherd) and which should be removed (background).

Tapping on the “Done” button (top right), the recogniser will be launched. This step requires Internet connectivity and takes the processed image and try to identify the decoration classes it belongs to. The recogniser returns 5 results ordered by a probability indicator (see the picture below).

If the user is satisfied by the results, may use the save button on the top right corner to save the results as a new sherd or inside an existing sherd.

Is important to note that only authenticated user can see the save button. Saving the results means telling the app in which site/assemblage the sherd should be saved and if we want to save it as a new sherd, which name assigns to it.

If the user is not satisfied by the automatically generated results may use the “Nothing good? Tap to add manually” button on the bottom right corner of the result window, to try a manual classification.

Essentially the user is able to visually compare the real potsherd with the images available in the catalogue (right-hand side in the image above) and when he/she finds a visual match, he/she uses the check button to save the selection. The images on the right may be changed swiping horizontally right-to-left (next image) or left-to-right (previous image).

Knowledge base 1.2

2019-07-25T00:40:41Z

The ArchAIDE knowledge-base is a collection of search & retrieval tools that allows accessing the database of reference classes for types, decoration and stamps. The defined classes derive from online and paper catalogues that were used to populate the “Reference Database”. A team of expert archaeologists and researchers are continuously working on the database extending the catalogue and refining the definitions.

When accessed from the “Tools” area of the dashboard, the knowledge base contains a brief description and the buttons to access the catalogues (see picture below). From the main menu, it is possible to access the specific catalogues directly.

Types

The search & retrieval interface for the type catalogue, shown in the image below, provide search field and filters to isolate types of interest.

The “keyword” search field (top of the form) is intended to apply to all the text fields of type records such as “name”, “distinctive features”, “period”, “origin”, “occurs”, “equivalents”. The filters are related to the main classification info such as:

Ceramic class (i.e. Amphorae, Terra sigillata, Maiolica di Montelupo, …)

Primary function (i.e. Transport, Food preparation and Consumption, …)

Form (i.e. Jar, Cup, Dish, etc.)

The result set is updated continuously while the user changes the values in the search form.

Tapping on one of the results in the list, the user will be directed to the detailed description of the selected type.

The detailed description of a type, see image below, contains all the relevant fields of the corresponding database record and the associated depictions.

Decorations

The search & retrieval interface for the decorations catalogue, shown in the image below, provide search field and filters to isolate decoration classes.

The “keyword” search field (top of the form) is intended to apply to all the text fields of type records such as “name”, “definition”, “period description”, “colour description”. The filters are related to the main classification information such as:

Decoration form (i.e. Painted, Incised, Burnished …)

Colours (i.e. White, Green, Yellow, …)

The result set is updated continuously while the user changes the values in the search form.

Tapping on one of the results in the list the user will be directed to the detailed description of the selected decoration class.

The detailed description of a decoration class, see image below, contains all the relevant fields of the corresponding database record and the associated depictions.

Stamps

The search & retrieval interface for the stamps catalogue, shown in the image below, provide search field and filters to isolate stamps of interest for the user.

The “keyword” search field (top of the form) is intended to apply to all the text fields of type records such as “name”, “definition”, “period description”, “workshop”.

In particular, the search engine allows to make searches including wildcard characters in order to match the text (for “Name stamps”) contained in the stamp itself. The scenario is the one in which the archaeologists find a sherd containing a stamp but only part of the name is readable or almost readable. To search a pattern like “ACCU …. R . S” (dots represents unreadable characters) the search engine allows to represent it as “ACCU%R_S” meaning “ACCU” followed by any character, then “R” followed by one character and finally “S”. Searching for this pattern allows identifying the stamp “ACCU ILARUS”. The wildcard characters that may be used in the keyword field are the following:

% for zero, one, or multiple unknown characters
_ for a single unknown character

To activate the pattern search on the stamp text, the first character of the pattern should be #.

Examples of patterns may be the following:

Typing "#a%" finds any values that start with "a"

Typing "#%a" finds any values that end with "a"

Typing "#%or%" finds any values that have "or" in any position

Typing "#_r%" finds any values that have "r" in the second position

Typing "#a_%_%" finds any values that start with "a" and are at least 3 characters in length

Typing "#a%o" finds any values that start with "a" and ends with "o"

The search is case-insensitive.

The filters are related to the main classification info such as:

Stamp type (“Name stamp” or “Pattern mark”)

The result set is updated continuously while the user changes the values in the search form.

Tapping on one of the results in the list the user will be directed to the detailed description of the selected stamp.

The detailed description of a stamp, see image below, contains all the relevant fields of the corresponding database record and the associated depictions.

Mobile App 1.3

2019-07-25T00:29:25Z

The ArchAIDE mobile App can be used on both smartphone and tablet on Android and iOS devices.

The App starts from a “splash screen” showing the project logo and the reference to the funding institution and program. The splash screen is the place where is carried on the synchronisation of the app with the ArchAIDE Server and thus all the related messages (i.e. “Loading …”, “Connection error …” etc.) will be shown.

The very first synchronisation may take a while (one to two minutes) because the whole Reference Database has to be downloaded from the server and initialised. The subsequent synchronisations are usually quite fast because it applies only to the changes occurred in the period between the first synchronisation and the subsequent.

From the splash screen, the user is automatically addressed to the Dashboard i.e. the home page of the App (see the picture below).

The dashboard was designed to support the user in his/her daily work providing quick access to the entities on which he/she is working, to the classification tools and to the users collected information (sites). The three areas composing the dashboard are:

Recent
This area contains the list of the entities (sherds, database records, etc.) visualised or edited by the user, ordered from the most recent to the oldest. Tapping on the entity will open the entity visualiser. The idea is that the user returning to the dashboard may access directly the entities he/she was working on the last time he/she accessed the app or a few minutes before.
Tools
In this area are shown the major tools provided by the App. The goal is to invite users to access and try them, in particular for newbies. Tapping on the tool icon, the tool interface will be open. The same may be achieved through the main menu.
My sites(authenticated users)Authenticated users will find here the list of their personal sites used to store sherd information. Using the “+” button the user will be directed to a form for the creation of a new site. The “swipe” action on a site button will show the “basket” button for removing the site itself. It is important to note that a site may be removed only if it doesn’t contain any sherd. Finally tapping on a site button, the Site viewer will be open.

Tips&Tricks(anonymous users)Anonymous users will see, in the right lower area of the dashboard, a message synthetically describing the app and the opportunity to register and become a known user.

The application main menu is positioned on the top-left corner of the dashboard as a “hamburger” button . The main menu allows to access the catalogues of the knowledge-base, the recognition and classification tools and the application settings.

Desktop application personal data 1.2

2019-07-25T00:08:48Z

Users that are already registered (via mobile app) may use the “Sign in” link on the top menu of the ArchAIDE Desktop site to log in and be enabled to his/her personal information.

Once the user logged in it will appear the menu item "Account" that allows to access the user personal sites and eventually logout.

The user's sites are represented as a grid of all defined sites with buttons to add a new site (upper right corner) e to remove a site (button right button in each site box) as shown below:

Clicking on the image or name is possibile to enter in the site and visualise the location info and access the assemblages contained. The "Published" info tells if the site is publicly available (i.e. open-data) or not.

The localisation of the site on the map is initially defined geo-coding the site address. The user may change the position dragging the marker around on the map.

Selecting one of the assemblages of a site the list of contained potsherds is shown (see picture below). Each sherd is described by essential data like ceramic class, form and (mandatory) identification code. More important are the classification information that tells to which type or decoration or stamp the sherd belongs to. The classification info may be added by the user manually or generated using the automatic classification tools and then saved on the sherd.

Each sherd may contain one or more depiction taken with the mobile device cam or with other technique. Each depiction can be used to start an automatic recognition process.

Personal data collected on mobile 1.1

2019-07-24T22:57:48Z

Only authenticated users may collect and store data through the app. The information about sherds should be registered under a specific assemblage that belongs to a site. All the registered users own at least one site (i.e. “Default site”) in which it exists at least one predefined assemblage (i.e. “Uncategorised”). Sites and assemblages helps localising the findings and are the minimum context information required to store a sherd.

Site

The Dashboard for authenticated users contains the list of user’s sites (My sites). Tapping on a site item, the site viewer will open showing the site address and the position on the map. The only required information about a site is the location represented as an address and coordinates.

To fix the geographical position of the site, “drag” (tap on the marker, hold, move) the marker around over the map. For an accurate positioning it is possible to use the zoom gestures to enlarge (zoom in) or restrict (zoom out) the map.

To edit the Site info (name and address) users would tap the edit button near the site name and access the editor window.

For each site, the user may define a list of assemblages containing the collected sherds. The only information required for the assemblage is the name. The assemblage is the container of the user collected potsherd.

Sherd

The collection of sherds composing an assemblage are presented as a grid showing the featured image of the sherd (the first added) and providing buttons to access the sherd details or remove the sherd (garbage basket). A new sherd may be added directly using the “+” button on the right bottom corner of the window.

Entering the details view of a sherd the user will be presented the following set of information:

Name and general info
Classification info
- Type classification
- Decoration classification
- Stamp classification
Depictions

The name and general information may be modified using the edit button (pen) near the sherd name. The classification info contains the list of classes (types, decoration and stamps) to which the sherd may belong to. The classification may be defined manually using the “+” button present in each tab or automatically using the classification tools. Using the swipe gesture on an element of the list is possible to remove the classification.

The depictions area contains the images collected for the specific sherd. In many cases, it may be useful to take a picture of the surface to get the decoration or the stamp or to take a picture emphasising the fracture in order to classify the form. The automatic classification tools may be accessed using the context menu associated with each image. Selecting one of the tools, the corresponding interface will be launched assuming the starting image as the initial parameter. At the end of the recognition process, the classification info may be saved on the sherd using the save button. The “share” item allows sharing the image with other people through the sharing tools available on the device (i.e. mail, WhatsApp, etc.).

Registered vs. Anonymous Users 1.2

2019-07-24T22:34:22Z

Users may register on the ArchAIDE Server and then be enabled to data collection services. Anonymous users may use: Knowledge-base (browse types, decorations and stamps), image recognition (without saving the results), shape recognition (without saving the results) and stamp search tools. The registered users may save potsherds and information about them on the server for future use.

Authentication and user registration may be accessed from the “user” button on the top right-hand side of the dashboard (left image below).

If the user is already registered, he enters username and password and then taps the Login button. On a successful login, the user will be redirected to the Dashboard.

After the first login, the user remains logged in even after closing and re-opening the application. To force the need for a new login, the user should log out (last item in the main menu) before closing the application.

Users that are not yet registered, using the “Register” button will access the registration form.

The registration of a new user requires only a few non-sensible data such as name, password, email, privacy. In order to activate the account, the system will send an email, to the submitted address, containing a validation link. Is important to check for the email even in the spam directory of your mail client because sometime it may happen that the automatic emails are classified as spam by servers. If you are waiting to receive the mail don't try to re-submit the registration form because your account is already registered. In case of problems contact ArchAIDE people through messenger or sending a mail.

FrontPage 1.5

2019-07-25T00:53:19Z

This Wiki area has the goal to help ArchAIDE users to use the mobile and web tools. Select the functionality you need support and follow the links. This version of the Help system is still under development and is not covering all the functionalities provided by the mobile (Android, iOS) and web tools but you will find information on all the most important.

Desktop Application