ImageMetry is a top-rated company developing custom mobile and web imaging products. Having a deep knowledge and experience in mobile and embedded devices, our solutions are often awarded by the most prestigious business and research authorities for their business as well as scientific impact.
To get ROI reports, business impact information or to see how we help our partners to get cutting-edge competitive advantage solutions, just drop us a line at firstname.lastname@example.org
Learn how we saved 2.7% of cost of a leading US insurance company
Download a report showing how our mobile OCR systems bring value to US and EU banks
Our face recognition solutions protect airports and borders in Asia
ONLINE AUCTIONS AND DATING
Learn more how in 2011 our apps increased revenues of our partners by 9.4%
Our solutions have been awarded by City of New York, Czech Government, research organizations, etc.
Face Recognition Across Illumination
The appearance of a face is dramatically altered by variations in illumination, facial expression, head pose, image size and quality, facial hair, cosmetics, accessories (such as eyeglasses), and age. To further compound the problem, we are often given only a few images of an individual from which to learn the distinguishing features, and then asked to recognize him in all possible situations.
In this project, we learned an exemplar-based classifier. Intuitively, we can render all kinds of variations for each person: all combinations of pose, illumination, expression, etc. this gives us possibility to recognize that individual under various situation.
Face Recognition Across Pose
In many face recognition tasks the pose of the probe and gallery images are different. In other cases multiple gallery or probe images may be available, each captured from a different pose. We have developed a face recognition algorithm that is able to handle large variations in pose using a “patch-based” rather than “holistic” representations.
General Face Detection and Recognition
Face detection and recognition refers to a technology in which human face is detected and a number of its features are recognized in arbitrary digital images automatically by a machine. The technology is primarily employed in images where human face is not the main object in the picture i.e. there are other objects in the image as well, the task of the algorithm is to recognize the face without any human interference at all.
You find our face technology algorithms is in various digital cameras with face detection feature. Here, we use mathematical algorithms to detect face and its features like a smile or eyes. Our further typical face related R&D projects projects are often used in biometric security systems to automatically detect and recognize faces of the people visiting an office building or a public place. Another widespread use of face recognition technology is in image database management by law enforcement agencies and other authorities. An image fed in the system is matched with the database and recognition is returned to the operator. The technology has evolved much since its advent a few years back, face detection and recognition algorithms can easily collect demographical information like age, gender, race and age etc. It is finding new areas of applications with every day passing. Moreover, so far, many large organizations have employed our R&D services in this field to automate the attendance recording process and keeping records of visitors.
The core of a face detection and recognition system is the algorithm used to detect the face. However, all face detection and recognition algorithms are NOT created equal; a less accurate method can give a large number of false results. When you shop for such an algorithm for your business or personal use, make sure you look for real professionals in the field.
Facial Expression Analysis
In this project, we delivered an automatic facial expression recognition system. Facial expression provides cues about emotion, regulates interpersonal behavior, and communicates psychopathology. Human-observer based methods for measuring facial expression are labor intensive, qualitative, and difficult to standardize across laboratories, clinical settings, and over time. To make feasible more rigorous, quantitative measurement of facial expression in diverse applications.
Gender and Age Recognition
Here, our algorithms take image of a person from and subjects it to analysis to extract information regarding the demographics of the person in the image. The technology founds numerous applications. Age and gender recognition technology is based special image processing algorithms, which enable the machine to recognize the person standing next to the camera as male or female without interference of a human.
Gender and age recognition technology will open new frontiers in the industry. As the technology enables a machine, which is most likely a computer, to differentiate the person based on demographics- the machine can then offer more specific services or messages to the person. Marketing firms are using web cameras to recognize age and gender of a buyer in electronics market and then based on the information they play targeted advertisements on TVs, which are of interest to the age group and gender.
Optical Character Recognition (OCR)
Our OCR development services provide convenience and automation, not only for the large organizations but also for the small enterprises and even individuals.
OCR softwares, typically, can read text from an image file and convert them into an editable and more manageable electronic file format. The technology has recently developed to a point where there is almost a zero percent chance of error.
Our partners use our printed or hand-written OCR development services, typically, for their projects in banks, mobile applications, insurance companies and idustrial visual inspection.
With an appropriate OCR program your bills, bank statements and many other documents can easily be converted into more manageable files. For large insurance businesses OCR can help automate time taking and labor intensive tasks in their process line.
Object detection and recognition
Object detection and recognition in images and videos is one of our most often provided services. To fulfill requirements of our clients, we use the most sophisticated algorithms. We are focused on PC platforms as well as mobile devices where the development strategy is totally different.
Our services in this area are often used in detecting cars, people, logos, OCR, re-captcha, weapons in hands of a person, etc. Furthermore, we help our partners often in detection of objects in large image databases which becomes particularly useful when the tags of the files are not according to their content.
ImageMetry’s object detection and recognition research&development services can be used in the field of generic objects as well as special ones.
Automatic license plate recognition
Automatic license plate recognition is a technology now widely spreading across roads and highways of many countries. The technology enables automation in many traffic related tasks. It is the key component in electronic toll collection systems on highways. ImageMetry’s automatic license plate reading solutions utilize a camera and a computer vision based OCR program which detects a license plate and then reads the characters on it. Our solution gives automatic precise license plate recognition in every condition and weather. Typically, we achieve accuracy around 99%.
So far, our R&D services in this area have been used by our partners in electronic toll collection systems, automatic vehicle identification, by law enforcement agencies, and at parking areas.
Image and Video Forensics
A digital photo is a representation of a discrete mathematical function having a finite set of digital values called pixels. When committing image forgery, the photo is opened in an image editing software (e.g., Adobe Photoshop) and modifications are carried out. These modifications change the algebraic and statistical properties of the image. What is important is that always they leave behind some detectable traces.
To determine whether a photo is original (not modified), our system analyzes it using a number of separate automatic methods. Here is a brief description of the core parts of the system. However sophisticated these core parts may be, we always provide the user with a simple to use report with clearly identified suspects.
Image and Video Ballistics
Found photos on a disk and need to know who (what specific camera) has taken those photos?
Found a set of photos and need to know if they have been taken by a same device?
Want to have a technology with an accuracy, in effect, 100%.
Using our camera ballistics methods, we identify whether a digital image or scanned document was taken by a particular device or not (by a particular person or not).
Optimization – Industrial visual inspection
We have developed automated inspection systems for manufacturers of all types of goods in all around the globe. Our machine vision solutions, typically, automate those steps of the process which requires a watchful eye. Cameras take visuals from the process line and submit them to central image processing program, which then analyzes them and reports a fault or a defected product; even better the system can remove the faulty products from going further in process automatically.
ImageMetry’s computer vision systems designed for inspection and quality control purpose generally consist of a single or multiple cameras (varies according to particular application) and a computing system which, based on its program, returns the desired output of the image. For more complex applications like inspection of complex irregular three-dimensional parts for faults, we use more specific and refined software programs. But, generally, consumer versions of our ready-to-use software packages are designed to be universal and flexible to reckon for different needs of customer.
Automating the inspection and quality control processes, in general,
- Save time by making the process much faster than manual inspection
- Help detect faults which slip away from human eyes like micro cracks and minute distortions in the object dimensions
- Measure the dimensions of an object with pinpoint precision
- Save cost by eliminating need of manual work and speeding up the whole process
Automatic Video Face Replacement
For our US client, we developed a method for automatically replacing facial performances in video. The method does not require substantial manual operation or complex acquisition hardware. We use a 3D multilinear model to track the facial performance in both videos. Using the corresponding 3D geometry, we warp the source to the target face and reatime the source to match the target performance. We then compute an optimal seam through the video volume that maintains temporal consistency in the final composite.
Accurate Camera Calibration from Planar Patterns
For our US client, we developed a novel camera calibration method by utilizing a single framework for square, circle, and ring planar calibration patterns.
Context-sensitive bicycle and pedestrian detection and tracking
We developed an algorithm to fuse sensor data from cameras for the purposes of detecting and tracking pedestrians and bicycles in proximity to autonomous cars.
Event Detection in Digital Videos
For a US client, we developed a system detecting a wide range of actions in video by correlating spatio-temporal shapes to over-segmented videos without background subtraction.