Category Archives: Terrestrial

Technology that can be used to record Archaeology and Cultural Heritage from the ground.

High Dynamic Range Photography/Photogrammetry – Part 1

High Dynamic Range (HDR) Photography

High Dynamic Range is a popular photographic technique that is used to produce more realistic photographic results or artistic images. It is a technique that can be used to try and replicate what the human eye can see as the dynamic range of a camera is limited and it is unable to record the lightest and darkest elements in a single photograph. This can be remedied by taking a number of photographs with varying shutter speed/aperture combinations and combining them using specialist software to produce a photograph with a greater dynamic range than can be recorded in a single photograph.

Many new digital cameras have the ability to produce HDR photographs using Auto Exposure Bracketing (AEB), special HDR settings (this may process the images for you resulting in an HDR photo on the camera but losing the originals) or manually setting the camera up.

Limitations of archaeological and cultural heritage photography

An intrinsic problem with taking photographs in archaeological and cultural heritage contexts is lighting; both too much and too little lighting are factors that hamper recording images that include as much detail as possible.

In the case of archaeological excavations attempts have been made to limit the problems in section photography by either reflecting more light in using white card or using a tarpaulin to cast a shadow. Both of these techniques work but require time and manpower.

The same problem is encountered in building recording where in an outside environment strong lighting can cause both bleached out areas and heavy shadows.

Netley Abbey - East Window

Netley Abbey, Hampshire – East Window. Photograph demonstrating the problem of strong oblique lighting causing both too much and too little light in the same image.

While lighting through windows can cause similar problems on the inside of buildings.

Netley Abbey, Hampshire - Sacristy/Library.

Netley Abbey, Hampshire – Sacristy/Library. Photograph demonstrating the effect of excessive light coming through a window causing the dual problems of both too much light near to the window and too little light in other areas.

In order to reveal elements in dark shadow a high exposure camera setting is required, while bright bleached out areas are only revealed with low exposures. These elements together with well lit areas cannot by revealed in one single photograph, this is where High Dynamic Range photography comes in.

HDR Photography in Archaeology

HDR Photography was introduced to Archaeology by David Wheatley in 2010, he provided examples of its use in improving the standard recording methods of excavations, cave sites and even using archived analogue archaeological photographs. Sadly its use was not embraced by the community probably due to technological limitations at the time and the inherent conservatism of the industry and museum archives which were yet to embrace digital photographic technology. Technology and the industry have now caught up with his ideas, with digital cameras being present on most if not all excavations, while other scholars have now begun to bring the technology to the technique of 3D recording using photogrammetry.

HDR Field Archaeology Photography

Photography is one of the primary recording techniques within field archaeology and has been since the introduction of discipline, but conservatism within field archaeology has meant that it was only fairly recently that digital photography became the primary recording technology.

Digital cameras have a number of benefits within archaeology:

  • The ability to take numerous photographs on one memory card.
  • No need to pay to process films.
  • No need to digitize the photographs.
  • Where once excavators may have been told to limit the number of photographs taken on an excavation to keep the processing costs down, digital media allows almost limitless photographs to be taken.
  • Photographs can be as easy as point and click with the camera controlling all of the settings.

But they also have drawbacks:

  • Where once archaeologists knew how to use an analogue camera to take bracketed shots, the automatic setting on digital cameras is commonly the only setting used as it produces results at a required level of quality, this means that the archaeologist may not know how to properly operate the camera.
  • Although almost limitless photographs can be taken, limits should be included as the archive may still need to be sorted through.
  • The requirements for digital storage can be complicated and costly.

Although not ideal, a number of modern cameras now come with an HDR setting on them which in many cases can be changed to the required level of bracketing, although only the merged photograph is saved losing the possibility of later re-processing the photographs with different settings.

Field Archaeology Archive Photographs

One benefit of traditional bracketing of analogue photographs for archaeological excavations is that they provide an ideal resource for conducting HDR processing. These archives have multiple photographs at different exposure levels which can be digitized and processed to provide better results than the originals and be re-entered into the archive with the digitized originals.

HDR using archive slides from excavations at the Cove, Avebury using archive slides (Wheatley 2010)

HDR using archive slides from excavations at the Cove, Avebury (Wheatley 2010)

HDR Building Photography

Building recording is an area that can be significantly enhanced by the use of HDR. It is difficult to provide adequate lighting in many cases, meaning that some areas are brightly illuminated while others are dimly lit loosing information in both cases.

Processing bracketed images into an HDR image provides a greatly enhanced image.

HDR Photogrammetry

Recent developments in camera technology, HDR software and photogrammetry software have allowed the introduction of HDR Photogrammetry. Thanks to the additional information present in the photographs models of higher detail and accuracy can be created in non-optimal lighting conditions.

As well as the ability to use tone mapped images produced from HDR images the Agisoft PhotoScan Photogrammetry software can also process .exr file format High Dynamic Range images into 3D models.

HDR Object Photogrammetry

One area under study is its use in photographing objects. The benefits are determined by the type of material used, some are greatly enhanced by HDR while others are little altered.

Image matching result from the images originated with different HDR processing: a) No HDR; b) tone mapped images from HDR processing (Guidi et al 2014)

Image matching result from the images originated
with different HDR processing: a) No HDR; b) tone mapped
images from HDR processing (Guidi et al 2014)

HDR Building Photogrammetry

We have already seen the benefits of HDR Photography in building recording and this can continue with photogrammetry.

Netley-Window-PG

Photogrammetry point cloud of the east window of Netley Abbey, Hampshire showing how the raking sunlight on the left-hand side of the window has bleached out the photographs and lost detail

Both the increased level of quality of the photograph and the higher amount of detail present in the 3D model can easily be seen in the HDR photogrammetry model.

Software Solutions

A number of software solutions are available for the processing of HDR photographs, these range from high end photographic software such as Adobe Photoshop and Lightroom, through to HDR specific pieces of software and even open source solutions. HDRSoft’s Photomatrix comes in a number of versions which include plugins for different software packages such as Adobe Lightroom, Photoshop Elements, Photoshop and Apple Aperture. With low cost solutions such as Fusion HDR or free open-source solutions such as LuminanceHDR also being available.

In order to be view-able on low contrast monitors and paper the images need to go through a process called tone mapping, this replicates the appearance of the high dynamic range photograph on these media.

Downloaded imaged can be batch processed in software such as Photomatrix setting up how many images need to be merged together with a number of preset or custom settings allowing the images to be processed exactly as required. These pieces of software can also compensate for slight movement between the recording of the multiple images. The resulting images can then be saved as either .hdr (Radiance) or .exr (OpenEXR) file formats which record the HDR information.

Batch processing of images within Photomatrix

Batch processing of images within Photomatrix

Benefits
HDR photography can record more information in both photographs and photogrammetry models. By using open Source HDR software it can be free. Many cameras allow multiple bracketed photographs to be be taken automatically only adding a few seconds to the recording process.

It is also possible in some of the software to open a folder full of images and have the software batch process it without any user intervention once the preset settings have been loaded.

Drawbacks
Among the drawbacks are the fact that as the camera is taking multiple photographs it is difficult to stabilize the camera by hand, otherwise there will be movement between the photographs. Although movement between photographs can be corrected if you are bracketing shots and using software the automatic HDR setting on the camera will probably result in a blurry image.

UAV HDR Photogrammetry

UAV HDR Photogrammetry is an area I will be studying in the future. It has great potential for recording but will require a careful balance of UAV hovering, a steady gimbal, fast shutter speed and an adequate depth of field. It will be discussed in a future blog.

Sources
Guidi, G., S. Gonizzi, and L. L. Micoli. “Image pre-processing for optimizing automated photogrammetry performances.” ISPRS Annals of The Photogrammetry, Remote Sensing and Spatial Information Sciences 2.5 (2014): 145-152.

Kontogianni, G., and A. Georgopoulos. “Investigating the effect of HDR images for the 3D documentation of cultural heritage.” World Cultural Heritage Conference 2014 – Euromed 2014 – International Conference on Cultural Heritage Documentation, Preservation and Protection. (2014)

Ntregkaa, A., A. Georgopoulosa, and M. Santana Quinterob. “Photogrammetric Exploitation of HDR Images for Cultural Heritage Documentation.” ISPRS Annals of the Photogrammetry, Remote Sensing and Spatial Information Sciences 5 (2013): W1.

Wheatley, D. “High dynamic range imaging for archaeological recording.” Journal of Archaeological Method and Theory 18, no. 3 (2011): 256-271.

Advertisements

Kula

Kula is a Kickstarter Project consisting of two lenses that allow Digital SLR cameras and smartphones to record in 3D.

The two systems use mirrors to record two images slightly offset horizontally from each other, the free software then transforms what is recorded for a number of 3D recording systems.

Kúla Deeper
The Kúla Deeper is a 3D lens designed for use with DSLR cameras, its screws onto the end of existing lenses. A number of adapter rings a available to enable connection to different camera systems.

It is currently availble for pre-order for $105.

Kúla Bebe
The Kúla Bebe is a 3D lens designed for smartphones.

Potential
They provide an easy and cheap way of recording in 3D using available technology with just the inclusion of the lens for the relevant device.

Limitations
Because the system records views from the left and rights in the same image this reduces the overall quality of the image as it is split in half for viewing. This also means that a landscape image is transformed into two portrait images limiting the field of view of the image.

Kodak PixPro SP360 Action Cam

The Kodak PixPro SP360 is an action camera aimed at the extreme sports recording market.

It is able to record 360º 10MP photographs and Full HD 1080p Video using its curved lens. Using the Pixpro Remote Viewer App it connects to other devices with iOS, Android or Windows operating systems using Wi-fi and NFC (Near Field Communication) allowing both control of the camera and viewing of the images/videos. The images/vidoes can be viewed in a number of different modes – ring, dome, front 180º and rear 180º, 360º panorama and globe. The system also has motion detector sensor which starts recording when motion is detected.

Much like the GoPro camera the system comes with a number of accessories for attaching it to different devices as well as a waterproof case which aare available in the different bundles.

The PixPro SP360 costs $349 for the Explorer bundle for entry-level and the Aqua bundle for watersports and $399 for the Extreme Accessories bundle.

Potential
The camera is able to record much more than a standard camera without the requirement for multiple cameras and the images/videos created can be viewed/exported in a number of different modes. It is water resistant, can be attached to a UAV or a person allowing a wide range of recording possibilities.

Limitations
Although the system can record a 360º view images/videos with the camera pointing upwards it is only able to record a 214º images/videos while pointing towards a subject. The distortion of a domed lens may also impact on the quality of the results.

GoPro HERO

The GoPro HERO is a new budget camera in the GoPro range. It takes 5MP (mega pixels) photos which can be recorded up to 5 frames a second and records video at 1080p30 and 720p60. With its rugged case it is waterproof to 40m.

The interface has been made easier with the new QuikCature mode which allows the user to power on the camera and start recording by pressing a single button. Pressing once records video, while holding the button down for 2 seconds starts to capture time lapse photos.

http://shop.gopro.com/EMEA/cameras/hero/CHDHA-301.html

The GoPro HERO costs £99.99.

Potential
The GoPro HERO has much of the same potential as the GoPro HERO4 which can be read here.

At only £100 its potential for budget wearable technology in archaeology is enhanced greatly, with the cost of 3 GoPro Heros being less than 1 GoPro Hero4.

Limitations
Up until now one of the major limitations of the GoPro has been the price, but with a GoPro costing £99.99 this drawback has been removed.

There are limitations to this budget model though, the camera is only 5MP as opposed to the GoPro HERO4 which is 12MP, which is quite a low level of recording for a modern digital camera. It will limit the amount of information recorded in a photograph, and hence any 3D models created from the results.

GoPro HERO4

The GoPro HERO4 is a new camera in the GoPro range, it consists of two separate models, the silver and the black.

Silver

The silver model includes a built-in touch-screen display, previously this had been available as an extra which connected to the back of the camera, the black model does not come with this.

It captures photographs with its 12MP camera at speeds up to 30fps and records video at 1080p60 and 720p120.

Black

Professional Video up to 4K30

It captures photographs with its 12MP camera at speeds up to 30fps and records video at 4K30, 2.7K50 and 1080p120.

Both models have built-In Wi-Fi + Bluetooth allowing connection and control from other devices using the GoPro App, Smart Remote2 and more.

With their rugged case they are waterproof to 40m.

The silver version of the camera is £289.99 while the back is £369.99.

Potential
The GoPro camera was designed with the extreme sports industry in mind, but it has since find a place in a number of other disciplines.

The size, weight and portability of the GoPro camera has meant that is had become the mainstay of the UAV (unmanned aerial vehicles) industry with the camera becoming standard on many quadcopters. The user can see what the camera sees by either using the on-board wi-fi and an app on tablet computers or phones; or a fpv (first person view) system which can be sent wirelessly to either a video monitor or video goggles. Motorised gimbals can be used to both, stablise the camera to remove any shake caused by the UAV, and to change the direction the camera is pointing in mid flight.

The GoPro also has great potential as wearable technology in the trench:

  • The camera comes with a number of mounts out of the box, including ones that can attach to hardhats, but other mounts are available including a chest harness. These could be used to record the excavation process by attaching GoPro camera to the archaeologist excavating; the video could be used to record video but the individual video frames, or photographs, could be used to create a 3D photogrammetry model of the excavation. The combination of head and chest mounted camera may provide enough images to make an accurate model of the excavation. The rugged plastic housing can protect the camera from anything that it comes across in the trench.
  • If a complicated or important excavation was being undertaken or an important artifact was being lifted the excavator could have a live conference with an expert who could advise them on the best way to excavate it.
  • Experimentation with aspects of wearable technology has been undertaken on the Portus Project in Italy under the auspices of the University of Southampton.

The use of the GoPro is not limited to a single camera, the Dual HERO System is a case that allows two GoPro camera to be put side by side which allows 3D footage to be created using the GoPro Studio editing software.

GoPro Dual HERO System

The 360heros series of GoPro mounts clips a number of cameras together allowing spherical video recording with up to 14 GoPro cameras, the videos recorded can be stitched together in video stitching software to create immersive 360˚ videos. Mounts are available for everything from the top of helmets, to UAVs, underwater recording and even for recording 360˚ 3D with 14 GoPro cameras.

The system can record immersive videos on the ground, in the air and underwater which can viewed in a video player online – http://video.360heros.com/website/users/webplayer/videodetails.php?vid=59

The Freedom360 provides a similar system.

A GoPro has even been attached to a shovel while archaeology is being undertaken – https://www.youtube.com/watch?v=ht0RR94O9jY&feature=youtu.be

It has been used in underwater test of sfm (structure from motion) – http://arc-team-open-research.blogspot.co.uk/2013/03/extreme-sfm-underwater-archaeology.html

Limitations
The HERO4 Black provides the user with a highly portable camera capable of recording 4K video, it does however come at a high price which limits the potential of using multiple cameras in an immersive recording system, the cheaper HERO does provide this opportunity however.

Microsoft Kinect for Windows v2

The Kinect for Windows v2 Sensor is the Windows equivalent of the Kinect which was shipped with the Xbox One Entertainment System. It is designed for the development of applications enabling humans to interact with their computers using gestures and speaking via the colour camera, infrared (IR) emitter and a microphone array built into the Kinect. The Kinect v1 for Windows was previously released after the Xbox 360.
http://www.microsoft.com/en-us/kinectforwindows/meetkinect/features.aspx
Kinect for Windows v2

The v2 has significant improvements over the v1, including three times higher depth fidelity, twenty times the resolution and the ability to record full 1080p video.
Kinect Features

But of importance to the archaeological and heritage recording community is the fact that the latest version of the Software Development Kit also comes with Kinect Fusion, a product that was previously available for the first Kinect. This software allows the Kinect to scan and record its environment in 3D thank to its 6 degrees of freedom transform (the original x, y ,z, roll, pitch and yaw of camera) which can determine the original position of the sensor and how much and in what way the kinect has moved between scans. This allows the Kinect software to align the original raw depth map (3D positional data recorded calculated from the time it takes for the light from the camera to reach the object and return (time-of-flight)) with all those that are recorded after it, which can then be used to create a mesh of the complete environment getting recorded, with a 3D representation being created. http://blogs.msdn.com/b/kinectforwindows/archive/2014/08/21/updated-preview-sdk-now-available.aspx

4201.kinect-for-windows-fusion.jpg-520x0

The Kinect for Windows v2 Sensor costs £159.

Potential
The previous version was used in a number of useful archaeological application such as;

As the Kinect v2 is much more powerful it has a much greater potential for being used as a cheap/quality object scanner. It also has great potential for creating interactive immersive site tours , possibly in combination with the Oculus Rift VR Headset, where the gestures of the user determine what happens in the 3D representation of the site.
It also has application in 3D GIS such as where studies in Archaeological Visibility could be conducted first hand within the environment.
If it can be made battery powered and a laptop with sufficient power is used it could also add to recording on archaeological sites, providing a quick easy method of 3D scanning objects/areas.

The Kinect has previously also been used in the design of autonomous UAVs, these scan the environment with the Kinect Sensor and determine where obstructions are and help to navigate, or what hasn’t been scanned and directs the UAV towards these areas.

Unity plugin.

Limitations
The hardware requirements of the Kinect v2 are quite high so providing it with a computer capable of running it may lesson any 3D scanning cost reduction presented by the cheap cost of the hardware itself. Microsoft provide a configuration tool to check whether a computer system is capable of running it.

  • Operating system – Windows 8, Windows 8.1, Windows Embedded Standard or Windows Embedded Standard 8.1
  • Processor – 64-bit (x64), physical dual-core 3.1-Ghz or faster processor.
  • Memory – 4GB RAM.
  • Graphics card – DirectX11 compatible graphics card.
  • Port – Dedicated USB 3 controller.

Similar Systems

The DPI-7 handheld mobile imager kit is a handheld scanner which attaches to a 7-inch tablet computer running the android operating system, it uses the Phi.3D real-time 3D capture software for mobile platforms.

Much like the Kinect it can perform realtime markerless camera tracking through its 6 degrees of freedom tracking and can create full coloured 3D models

http://www.dotproduct3d.com/Phi3D.php

The DPI-7 costs $5,499, which is significantly more than the Kinect, although it does simply interface with an Android tablet making it portable straight away.

dpi7_1t

The Structure Sensor Kickstarter Project by Occipital is a device that attaches to mobile devices, only the iPad currently, and using structured light the infrared sensor creates a depth map of the scene it is recording, the colour data is gathered from the camera of the iPad itself.
Occipital

Interestingly, PrimeSense the company who developed the infrared detector for the Structural Sensor and the 3D sensor for the first Microsoft Kinect, has recently been purchased by Apple suggesting that they will integrate the technology into their devices – http://appleinsider.com/articles/14/07/11/apples-secret-plans-for-primesense-3d-tech-hinted-at-by-new-itseez3d-ipad-app
A recently released App in the App Store, itSeez3D, also uses the Structure Sensor its 3D Scanner. The itSeez app can upload its captured data into the cloud for advance processing, with the resultant 3D model being delivered within minutes.

http://appleinsider.com/articles/14/07/11/apples-secret-plans-for-primesense-3d-tech-hinted-at-by-new-itseez3d-ipad-app
http://uk.pcmag.com/occipital-structure-sensor/35223/review/occipital-structure-sensor
http://www.theguardian.com/technology/2013/nov/25/why-did-apple-buy-primesense-for-a-key-technology-itll-deploy-within-a-year

The sensor costs $379 for the sensor alone, or $499 together with the Skanect Pro 3D Scanning software. You need to purchase an iPad as well to attach the sensor to.

Both of these technologies provide a completely mobile method of recording.

Raspberry Pi Model B+

The Raspberry Pi Model B+ is the latest model in the Raspberry Pi stable of computers. It is a credit card sized computer that can run a number of operating systems including Linux and Android, it was originally aimed at teaching children computing, programming and electronics but has been embraced by a whole community of people interested in its potential.

Raspberry Pi Model B+

With the GPIO (general purpose input/output) pins on the board many different electrical devices can be controlled; from turning an LED (light emitting diode) on and off to driving motors and taking readings from distance measuring devices. This makes it very useful for controlling robotics such as UAVs (Unmanned Aerial Vehicles) and Rovers.

By the inclusion of a battery pack it can be made completely portable, with the computer being controlled remotely via Wi-Fi using a wireless dongle and using either the SSH (Secure Shell) command prompt interface or the VNC (Virtual Network Computing) virtual desktop application, you can even use your mobile phone – http://www.raspberrypi.org/piui-control-your-pi-with-your-phone/.

The model B+ costs around £25.

Camera Module
Although the Raspberry Pi itself comes with no digital recording capabilities out of the box, for another £25 the camera module can be purchased and used to record video at 1080p30, 720p60 and VGA90 modes and images can be captured with its 5 megapixel camera.

Potential
The previous versions of the Raspberry Pi have already been used in robotics projects such as controlling Rovers and UAVs to controlling photographic recording stations. With extra GPIO pins and a more powerful computer this potential can be added to.
http://makezine.com/2013/07/06/the-raspberry-rover/
Rover

http://www.botched.co.uk/
UAV

It can be used to make cheaper versions of technology which already perform a usefully function; such as this gigapixel camera rig which takes overlapping DSLR photographs with preset motorised rotation and elevation of the rig set by the software on the Raspberry Pi, with the rotation being also controlled by the Raspberry Pi using a stepper motor whose rotation can be set accurately. The images can later be stitched together in software. A commercial version of this system is the GigaPan EPIC Pro gigapixel camera rig which costs $995. The software for the project is free to download. Free software solutions for stiching panorama photos together also exist, such as http://hugin.sourceforge.net/ and Microsoft’s ICE (Image Composite Editor) http://research.microsoft.com/en-us/um/redmond/groups/ivm/ICE/
Gigapixel Camera Rig
An example gigapixel image using the rig can be seen here.
Gigapixel
An example of a GigaPan capture undertaken with an EPIC Pro by the author and a colleague on the site of Netley Abbey in Hampshire as part of a recording project can be seen here.

Netley Abbey - GigaPan

Maker movement companies such have Adafruit have embraced the Raspberry Pi designing by building and selling many addons, including motor controller boards and touch screens.

Limiting factors
The Raspberry Pi can run many useful pieces of software but it is limited by it’s processing power, although multiple Pi’s could be used for different parts of a project.