CCPi (Tomographic Imaging) Case Study

Tomographic Imaging of a Nokia 702 Mobile Phone and other Items

The Challenge

Spot anomalies using the software available. As part of an assignment a set of objects were scanned using an X-Tech 320 kV CT scanner. This was followed by an exploratory stage using the multiple methods available from visualisation systems.

Key items to be discovered included the following:

  • Understand how different materials and components can be separated.
  • Extract some understanding of the objects to gain insight in to the object’s function
  • Extract the shape of a visible or unknown component.
  • Analyse the density of the materials used and identify defects.

The Solution

Presented a range of 2D and 3D views that could be manipulated to see features. An ancient mobile phone, the Nokia 702 amongst other items were investigated. About 1500 2kx2k X-ray images were captured over evenly spaced rotation angles of the mobile phone. These were reconstructed to create a 3D volume approx 1500 x 1000 x 1000.

Example 1: Mobile Phone Nokia 702

The speaker mechanism (bottom) can be clearly seen, as well as loose components (right) that were not screwed down properly.

Left: Speaker componets clearly visable idefining 3D shape and measurable size and density. Right: Loose componets (extra screw) that has fallen into the phone during assembly is visable and measurable.

Left: Speaker components clearly visible identifying 3D shape and measurable size and density.
Right: Loose components (extra screw) that has fallen into the phone during assembly is visible and measurable.

The obvious aerial is actually an aesthetic feature, as it is not connected, with a mesh being used instead (below).

Two density transfer functions highlighting the mesh aerial.

Two density transfer functions highlighting the mesh aerial.

Very quick confirmation of discoveries has occurred but the full visualisation component can be seen in the following screenshot that describes and views the raw data, image projections, 3D reconstruction slices and final 3D volume visualisation (left to right).

Mobile Phone being viewed as raw data, image projections, 3D reconstruction and final 3D volume visualisation (Drishti)

Mobile Phone being viewed as raw data, image projections, 3D reconstruction and final 3D volume visualisation (Drishti)

Example 2: Golf Balls.

A Fitleist 2 (High-quality) versus a Srixon 4 (standard quality) golf balls; were both scanned at 2kx2k and then investigated using identical networks to discover material characteristics non-destructively. The high-quality golf ball has a very well-defined liquid core with an injection point, multiple tight winding impressions, as well as, far more complex layers of materials all visible and can be quantified in terms of shapen and size.

gb12

Example 3: Low Quality Padlock.

A very low quality padlock – with simple mechanism. Although the padlock uses a key the shape has been shown to be irrelevant to opening the lock. All that is needed is for the clip (far right) to be pulled apart so virtually any flat shape will work.

Visualisation network for the low-quality padlock. Although simple can still with interactive exploration discover new features within the volume data set.

Visualisation network for the low-quality padlock. Although simple can still with interactive exploration discover new features within the volume data set.

Example 4: High Quality Padlock.

A very high quality spherical security padlock – with complex mechanism. Requires a complex key structure, but it is still possible to retrieve the main components of the key shape.

Various views and colourmaps highlighting components within the high-quality padlock; as well as marked up components showing a possible 2D key structure to passist in picking the lock.

Various views and colourmaps highlighting components within the high-quality padlock; as well as marked up components showing a possible 2D key structure to assist in picking the lock.

Example 5 Homeland Security combination padlock

Homeland Security combination padlock. Understand how both combination and key mechanisms work; 1. Extract the correct combination code for unlocking the padlock, 2. Extract the shape of the key, and 3. Analyse the density of the materials used within the padlock manufacture and identify defects. Visualization filters can extract specific components – in this case recover non-destructively the combination lock values for the padlock. Alternative visualization filters allow analysis and measurement of defects introduced during the manufacturing process.

 

How to extract the combinations from a padlock. Step-0by-step process extracting and examining the components.

How to extract the combinations from a padlock. Step-0by-step process extracting and examining the components.

Density measurements illustrating poor and good quality manufacture wiithin the lock. Network is complete network for the extraction of the combinations.

Density measurements illustrating poor and good quality manufacture within the lock. Network is complete network for the extraction of the combinations.

 

The Benefits

The benefits of using visualisation in this case was to gain Insight as the human was able to see anomalies that an automated computer system could not. The question is when would an automated system – say for defect detection – be sufficient and used and then visualisation is not needed just data analysis and mining.

Credit to the students and researchers in the past who have created these visualisations and carried out the analysis.

Advertisements

Impact of Visualisation – a formula

Can visualisation methods be evaluated and should they.

A proposal presented at EuroVis 2014 held in the University of Swansea, Wales.

V = -T + I + E + C

This is the Value of a Visualisation is equal to the Time taken to understand the visualisation from the user plus the Insight the user gains from this plus the Essence gained for parts of this plus the Clarity defining the overall acceptance of the global data set by the user.

So let’s try this in action on a simple 3D example with users.

Example 1:

LiDAR visualisation for the geology community - can show rock escarpments and cave structures at myltiple sclaes with augmented meta-data and markers.

LiDAR visualisation for the geology community – can show rock escarpments and cave structures at myltiple sclaes with augmented meta-data and markers.

The previous image shows a geological structure in Egypt, Mount Sanai in stereoscopic 3D projection mode, with added markers showing way-points, landmarks, measurement fields as well as allowing ares of curvature for example to be highlighted and measured.

The time taken to do a demonstration is very low, and the number of Insight points (sometimes called Impact or Wow elements) is reasonably high as the user can discover anomalies, for example smooth areas of curvature in the rock face showing different strata. The Essence or global structure and understanding is also high as this is an intuitive 3D structure, and from this Confidence or Clarity in the data can be achieved. There are a few outliers and anomalies from the LiDAR data but the visualisation is very clean.

 Example 2:

CCP4software demonstration visualisation image - showing switchable components from proteins viewed in stereosc

CCP4software demonstration visualisation image – showing switchable components from proteins viewed in stereoscopic mode to an audience.

 

 

Roles of Collaborative Computational Projects

What are CCPs

“The Collaborative Computational Projects (CCPs) are networks that bring together leading UK expertise in key fields of computational research to tackle large-scale scientific software development, maintenance and distribution. They typically include academic, industrial and government representatives. Mature projects represent many years of intellectual and financial investment. The aim is to capitalise on this investment by encouraging widespread and long term use of the software, and by fostering new initiatives such as High End Computing consortia. In new project areas, there will normally still have been a significant amount of prior effort dedicated to producing software and a demonstrated community of users.” (from RCUK description)

What do CCPs do?

“The CCPs enrich UK computational science and engineering research in various ways. They provide a software infrastructure on which important individual research projects can be built. They support both the R&D and exploitation phases of computational research projects. They ensure the development of software which makes optimum use of the whole range of hardware available to the scientific community, from the desktop to the most powerful national and international supercomputing facilities. The training activities of CCPs have been outstandingly successful, benefiting several hundred students and post-doctorates each year.”

Need for Visualisation

Open question on how many of these need to promote themselves and what tools do they use to describe their data. Been asked to survey and look at passively and in case actively the use of visualisation tools within these and related communities.

 

 

Intro to the Visualisation Group at STFC

The Visualisation Group, part of the Technology Division within SCD, has been founded to support and maintain visualisation software and skills for large projects and user communities across STFC. We are about to launch a new website and some experiences will be relayed informally.

People involved in this work are: Martin Turner, Terry Hewitt, Rob Allan, Sri Nagella, Ron Fowler and Barry Searle. We also work closely with members of the Virtual Engineering Centre, Lynn Dwyer, Nando Milella and Iain Cant; and other groups across SCD. If you wish to visit (based in the North west location of Daresbury and in Oxfordsire in Harwell) or enquire about using the facilities then please email: hartree@stfc.ac.uk

 

This includes working with the Hartee Centre hosting their visualisation centres connecting outputs from some of the largest computer in the UK making them human understandable; with the TSB Space Catapult and ESA producing bespoke solutions for their data analytical and command and control needs; creating remote and distance data gathering and visualisation workflows to control the computational processes and providing specialist local high-end equipment within the centres that are near to the main STFC image capturing x-ray, neutron and laser facilities.

The group is working to support the high-end visualisation centres within STFC, with the key objective to consider the human-in-the-loop as an integral part to pre- mid- and post-data visualisation from the major facilities, from archived data stores and from computational simulations. This it is believed is a key component to increasing the efficiency of the major STFC facilities allowing researchers’ work-plans to be controlled, changed and even stopped on the fly.

The group has an emphasis to work in harmony with collaborators across the STFC mission. To this end there are links to strategic partners; as well as the Hartree Centre this includes the Virtual Engineering Centre (University of Liverpool centre based in the Daresbury Labs.), the Harwell Imaging Partnership (based in the RAL campus), the Collaborative Computational Projects (CCPs based across SCD) and a range of projects with other teams and groups within SCD and STFC.

 


 

One of the emphasis is on supporting local and national industry, to see how large screens and visualisation can link to their work.

3D Giaia software flythroughs for educational purposes. Use of 3D visualisation to aid the learning process.

3D Giaia software flythroughs for educational purposes. Use of 3D visualisation to aid the learning process.

Visualisation Does Matter

This blog allows for discussion and pointers as well as some images – that demonstrate uses of visualization from the UK. It is a personal travel by the authors and editors, and as such does not endorse any of the links to come.

Martin

£d Tomography visualisation at Atlas Visualisation Centre. Mark Basham from Daiamond Light Source

3d Tomography visualisation at Atlas Visualisation Centre. Mark Basham from Diamond Light Source (RAL)