Do you need to know if a digital image came from a particular digital device?
Do you need to know if a particular digital device created a particular digital image?
Do you have a set (large or small) of digital images and need to know if any of the set are related to each other because they were taken by the same device?
Forensic Image Analyser can help!
Forensic Image Analyser (FIA) speeds up investigations and massively reduces cost. Forensic Image Analyser (FIA) can test thousands of images at the same time and discover if any of them were taken by the same device. If you have a camera, mobile phone or scanner you can discover if any of the images were taken by your suspect device or not. If you have a database of images Forensic Image Analyser (FIA) can process these regularly to test for links.
- Forensic Image Analyser (FIA) uses ‘Sensor Pattern Noise’ (SPN) created by the silicon chip and embedded in the image as a ‘fingerprint’. Every silicon chip creates a different SPN and so each digital device has a unique identifying feature. Natural imperfections in the silicon chip and different sensitivity of pixels to light.
- The uniqueness of the imperfections makes SPN a robust fingerprint.
- SPN can differentiate devices of the same model.
What Devices Does Forensic Image Analyser (FIA) Work With?
Forensic Imge Analyser (FIA) will work with all digital cameras worldwide and 75% of Mobile Phones which include the latest HTC’s, iPhone, Blackberry and even iPad images.
Why not give Ben a call to discuss a demo on +44 (0)1827 255 170 or Clint on +61 428583545. Alternatively use the contact form and we will get back to you.
What Can Forensic Image Analyser (FIA) Be Used For?
- Intellectual Property Protection and Copyright Protection;
- Proof of Ownership and Audit Trail of images;
- Criminal and Terrorist Investigations;
- Paedophile, child pornography, and child protection investigations;
- Proof of sharing habits and patterns of behaviour for commercial purposes;
- Sharing habits and patterns of behaviour for crime investigation and terrorist purposes;
- Detection of networks of sharing images – commercial and legal (police and security)
How does Forensic Image Analyser (FIA) do it?
The SPN, n, which appears as a high frequency signal, can be extracted from an image, I, n = I – F (I) eq(1) where F is a denoising function.
The SPN obtained from eq (1) can be improved by using the enhancer to attenuate the interference of scene details. There are several enhancement models that have been developed and the choice of a model depends on the strength of the SPN components in the image, but all the models attenuate the stronger signals in the image to allow the SPN to be maximised. The enhanced SPN increases the identification rate and allows the use of smaller crops of images to be used for classification. The higher the magnitude of the scene details is, the enhancer applies stronger attenuation on the details.
Although it has been known for some time that images possess a signal generated during the imaging process it has not been possible to use it due the weakness of the signal. The weakness in the signal meant that it would often result in a misclassification therefore the use of the signal for these purposes was fruitless. Forensic Pathways has developed a method to overcome this and enhance or amplify the signal to a point that it can be measured reliably and used effectively. Prior to this the method was unknown globally and hence the need for the International Patent Application we have submitted.
Our technologies are innovative because they introduce two new methods namely (1) Sensor Pattern Noise (SPN) enhancement and (2) Unsupervised Classification.
The current technique for extracting the sensor pattern noise or digital fingerprint from an image suffers from high misclassification rates because the fingerprint data is contaminated by the scene detail in the picture. The innovative step introduced in Forensic Image Analyser (FIA) is to enhance the SPN or digital fingerprint after it has been extracted to massively increase the rate of accurate identification and classification.
What Are The Key Benefits of Forensic Image Analyser (FIA)?
- Automatic batch processing of digital images
- Identifies if an image or a number of images were taken by a suspected device
- Identifies those images in a set taken by the same device or not
- Saves time and money due to multiple images being processed
- Speeds up investigation by identifying source devices and linked images
- Processes images from any storage media (laptops, pen drives, cameras, mobile phones etc.)
Case Study 1
You have found a set of 1,000 images on a Laptop Computer and recovered two cameras. You need to know if any of these images were taken by the cameras you have recovered.
The images and cameras can be compared and we can tell you if any of them are linked. This will work with large numbers of images and large numbers of cameras.
Case Study 2
You have found a set of 1,000 images on a Laptop Computer but no cameras. You need to know if any of the images have been taken by the same camera. The images can be compared and we can tell you if any of them were taken by the same or different cameras.
Case Study 3
You have a large library, collection or database of images recovered from different places over a long period of time. You need to know which of these images originated from the same camera.
Whether you have a set of suspect cameras or not we can tell you which of the images in the collection are linked to the same source going back any period of time.
Currently Forensic Image Analyser (FIA) is being offered as a service. There are two options to choose from listed below. Please Contact Us to find out about pricing and image handling.
Image Identifier – Matching Images with Source Devices (Camera is available for comparison)
This technology is used to identify which of a set of images have been taken by which Device (camera). Typically, the Investigator will have recovered a number of images and a number of devices (cameras etc) and wants to know which device created which of the images. In this situation the investigator will have possession of one of more suspect devices (cameras) to test against the images.
Classifier – Matching Images with Other Images (No Camera is available for comparison)
This technology clusters a large set of images taken by an unknown number of known devices (cameras) such that each cluster contains the images taken by the same device. The investigator in this case will only possess a set of images BUT NO DEVICE (camera) for comparison and matching. Instead, the images are matched against other images and without a camera being available. This technique is simply matching images with other images. In this case the investigator will typically want to know which of the images were created by the same device (camera) even without the camera having been recovered.
These technologies deliver the following new functionality:-
- The ability to prove whether an image was taken by a particular camera or no
- To prove that an image possesses the same digital signature as another image & hence they are from same source
- To prove that an image is owned by a particular person due to the evidence of originality presented by their ownership of the camera
- To prove that an illegal image in the possession of a person was or was not taken by their own camera
- To prove the innocence of a person alleged to have taken an illegal image with their own camera when the signature demonstrates that another camera took the image
- To prove that an image infringes the digital rights of another individual by showing that the image possesses the same signature as a protected image in a collection owned by a third party
- Prove breach of Copyright or Intellectual Property violations of an image
- To prove relationships between individuals based on the sharing of images bearing the same digital signature
- To prove relationships and patterns in the use of device cameras in use around the world based on recovered images which bear the same digital signature and hence demonstrate patterns of use (for example paedophile networks
- The technology will be used in future to present opportunities for mass screening of images and automated classification
- The technology will be used in the future as a new form of indelible image ‘water-marking’. This will enable individuals to mark images secretly with a marker to show ownership. Water-marking is currently undertaken but it involves planting a piece of code into the image. Our technology will use the signature itself located in a secret location in the image. The signature will prove the origin and the ownership.
Forensic Image Analyser (FIA) and the techniques it uses are protected by PCT Patent Application Number PCT/GB2010/050247.