News Letter, object recognition and tracking
0 likes421 views
The document discusses the advancements in real-time object recognition and tracking technologies developed by Concurrent Vision APS to enhance automation in robotics and various applications such as video surveillance and biomedical image analysis. Utilizing digital ASIC and FPGA technologies, these systems enable true real-time processing of high-speed moving objects and improve accuracy in settings like active surveillance and medical diagnostics. Additionally, the document highlights the application of computer vision in assisting the visually impaired through innovative wearable devices.
News Letter, object recognition and tracking
- 1. February 2010 | Concurrent Vision ApS | +4530328964 1.issue, February 2010 Real Time Object Recognition and Tracking Concurrent Vision ApS Automation systems, especially in the world of robotics, are becoming faster creating an increasing need to track objects at higher speeds than ever before. Systems which rely on computer vision analysis to make artificial intelligence decisions and provide control, extend from high speed production lines and robot arms to autonomous guided vehicles, missiles and planes. Such systems use computer vision algorithms to extract information from images in a video sequence to identify and track objects in a scene. Usually these algorithms require high computational resources from a general purpose processor or a DSP, causing high computational latencies. High latencies act as a prohibitive factor for providing true, real time recognition and tracking of objects moving at high velocities. Company Concurrent Vision ApS develops real time, high speed, vision-based systems that identify and track objects in a continuous video stream. These systems are based on the digital ASIC and FPGA technologies to implement high speed parallel computations providing true real time recognition and tracking of objects moving at speeds above 200 km/h. Typical applications of these systems include active video surveillance, vision- based robotic arms motion control, providing cognitive characteristics to robots and tracking high speed moving targets. Of other applications can be mentioned video stabilizing, augmented reality, image stitching, real time demosaicing for high definition video cameras, 3D imaging, intelligent toy and physical interactive computer games. Concurrent Vision also provides solutions for the acceleration of high speed content based image retrieval systems that search for digital images in large databases. An example of such systems is retrieval and matching medical images for computer aided diagnosis. Contents Intelligent active video surveillance Biomedical Image Analysis Visual Feedback control of robots Diverse Object recognition applications
- 2. 2 Concurrent Vision ApS Intelligent Active Video Surveillance STATE-OF-THE-ART COMPONENT TECHNOLOGIES IN VIDEO ANALYSIS FOR SURVEILLANCE Automated video surveillance in commercial, law Concurrent Vision, In enforcement, and military applications is concerned with key areas of real-time observation of people and vehicles in crowded Video-based detection and environments. A type of observation that tends to describe tracking, actions and interactions and probably predict behavior. Video-based person Active surveillance as a real-time medium creates effective identification, and deterrence systems protecting people and businesses from Large-scale surveillance crime and criminal activity. In continuous automated systems. monitoring of surveillance video, security alerts are issued responding to burglary in progress or to suspicious individuals, moves or objects in a scene. Automated Video surveillance technology has also been proposed in applications to measure traffic flow, detect accidents on highways and log routine maintenance tasks at nuclear facilities. Military applications include patrolling national borders, measuring the flow of refugees in troubled areas, monitoring peace treaties, and providing secure perimeters around bases and embassies. Such video surveillance presents a number of technical issues including moving object detection and tracking, object classification, human motion analysis, and activity understanding. Concurrent Vision’s solutions solve or aid solving technical issues of Automated Video surveillance by providing high speed techniques for the following: Detection and tracking which involves real-time extraction of moving objects from video and continuous tracking over time to form persistent object trajectories. Human motion analysis which is concerned with detecting periodic motion signifying a human gait and acquiring descriptions of human body pose over time. Activity analysis deals with parsing temporal sequences of object observations to produce high-level descriptions of agent actions and multiagent interactions.
- 3. 3 Biomedical Image Analysis Computer Aided Diagnosis Reduces Human Errors Matching Medical Images for Computer Aided Diagnosis Medical databases contain huge amount of information relevant to illnesses Some studies show that 20 to 40 percent of statements made on radiological and their cures. These databases contain radiologic medical images that give reports by radiologists or radiology residents were pictures of small details of organs in the body. Benefiting from these images is found to be erroneous. however quite difficult, since data sets to be analyzed by radiologists is Errors can be classified as increasing substantially. Automatic image retrieval and matching Systems observational and based on scale and rotation invariant object recognition techniques, can be interpretational errors. used to collect or classify the statistical information obtained from the Observational errors can be databases, and perform computer aided diagnosis of diseases and linked to incomplete or abnormalities. Computer aided diagnosis improve accuracy of statements made faulty search patterns. Observation is for instance on radiological reports and reduce both observational and interpretational enhanced by taking human errors in these statements. Automatic retrieval and matching of medical advantage of the computer images takes advantage of the merging of medical imaging with multimedia ability to see shades of technology in networked multimedia systems for image-assisted medical care. gray beyond the range of Object recognition techniques depend greatly on extracting and detecting features in human vision and ability to 2-D scalar images. Feature points are used to establish correspondence between pairs use sophisticated search of images which is important for landmark based image registration and for building patterns. Computers can statistical models of shape and appearance. Extracting and matching Features in store and analyze all 1000 images can for instance be used in content based image retrieval from a database of shades of gray in the fracture images for the purpose of planning surgical interventions after fractures. photon beam exiting the Image retrieval and matching can be used to supply similar cases to an example to patient during radiologic help treatment planning and find the most appropriate technique for a surgical scans. Shades representing intervention. The Figure below shows examples images of a fracture database used for differences in bone and retrieval. tissue density, whereas Human visual range can only see 32 or fewer shades of gray. Errors of interpretation can be linked to the practitioner’s failure to link abnormal radiologic signs to relevant clinical data. Using object recognition, image retrieval and matching Concepts used in the computer vision technique for extracting and matching features algorithms, computers can in 2D scalar images can be extended to scalar images of arbitrary dimensionality. access and process huge Retrieval and matching of 3D human Magnetic Resonance Imaging (MRI) brain scans amounts of stored clinical and 4D computed tomography (CT) cardiac scans are examples. data and produce accurate interpretations
- 4. 4 Continuous real- time monitoring of Object Recognition in Multimodal vasospasm using Biomedical Imaging TCD Cerebral aneurysm refers to Extracting and matching features for the localized dilation or object recognition can be exercised to all ballooning of the cerebral types of medical images acquired by any artery due to the weakening of the wall of the blood existing image acquisition modality. It can vessel. As the size of the for example be used to automatically aneurysm grows, the detect and diagnose knee meniscus tears chance of it to rupture from MR medical images. It can also be increases. Rupture of the used to perform real-time analysis of cerebral aneurysm will lead to subarachnoid Transcaranial Doppler Ultrasound hemorrhage (SAH), which (TCD) image streams for such purposes is a serious condition with a as the study of cerebrovascular ischemia mortality rate of 30-60%. (stroke), the monitoring of blood flow The primary treatment for velocity during intensive care, general this condition includes open anesthesia and carotid endarterectomy surgery aneurysm clipping and endovascular coiling. (CEA), the detection of vasospasms after Regardless of the subarachnoid hemorrhage (SAH) and the treatment, patients suffering assessment of arteriovenous malformations (AVM).For instance, many morphological from SAH may undergo and dynamic properties of the common carotid artery (CCA), e.g. lumen diameter, vasospasm, which is a distension and wall thickness, can be measured non-invasively with ultrasound (US) condition when blood techniques. This however requires as a preliminary step the manual recognition of the vessels spasm, leading to decreased oxygen delivery. artery of interest within the ultrasound image. In real-time US imaging, such manual It is most likely to occur initialization procedure interferes with the difficult task of the sonographer to select within 3-7 days after and maintain a proper image scan plane. Even for off-line US segmentation, the treatment. As a result, requirement for human supervision and interaction precludes full automation to continuous monitoring of eliminate user interference and to speed up processing for both real-time and off-line the blood vessels within the applications. first 3-14 days of after SAH is desired to assess the Automatic object recognition and tracking can also be extremely useful in conjunction presence of vasospasm. At with Optical Coherence Tomography (OCT). OCT is an imaging technique that the present time, there are allows non-invasive, high resolution, cross sectional-imaging of both transparent and various accepted clinical non-transparent structures. The greatest advantage of OCT is its resolution. Standard methods to diagnose resolution OCT can achieve axial resolution of 10-15 µm. A high resolution OCT vasospasm. A non-invasive technique to monitor increases the resolution to the sub-cellular level of 1-2 µm. Below, a figure showing a vasospasm includes the true subcellular image using OCT with a resolution of 4 µm use of Transcranial Doppler Ultrasound (TCD), which is cost-effective, easy to use and potentially available 24- 7. TCD is a tool that transmits ultrasound to measure the blood flow velocity in the blood vessels, which acts as an indicator for the occurrence of vasospasm. However, the use of TCD requires the presence of a skilled OCT has demonstrated feasibility for high-resolution imaging of the vascular system ultrasonographer, and and other vulnerable tissue. This includes the central nervous system and the cartilage suffers from operator of joints. OCT can be applied to a variety of applications such as dependence. The use of Diagnosing and monitoring of retinal diseases computerized monitoring Imaging atherosclerotic plaque improve sthe current TCD technology and minimizes Tumor detection in gastrointestinal, urinary, and respiratory tracts the need of dedicated Detection of skin Cancer ultrasonographer. Early detection of osteoarthritic changes in cartilage
- 5. Real-time in vivo OCT is typically used for assessing arterial wall pathology in vivo. It may provide more Brain Tumor detailed structural information than other techniques. With high resolution OCT, and automatic object recognition, atherosclerotic plaques can be diagnosed in real-time Microvasculature with high accuracy including measurement of the thickness of thin fibrous caps less Monitoring Using than 65µm. This represents a step towards in vivo assessment of the risk of rupture. Combined Laser Insight into the physiology of a plaque is complementary to the structural information Scanning Confocal offered by the OCT grayscale image. While the OCT image presents morphological Fluorescence information in highly resolved detail, it relies on interpretation of the images by trained Microscopy and readers for the identification of vessel wall components and tissue type. Computerized Optical Coherence image retrieval and matching as well as object recognition can help the interpretation and identification process. It can be used to characterize different atherosclerotic Tomography in plaque components by their distinctive signal patterns as shown in the next figure. The Preclinical Figure shows histopathologic (hematoxylin and eosin staining; magnification ×40) and Window-chamber OCT images of a predominantly lipid-rich plaque in quadrants I—III. Models Glioblastoma multiforme (GBM) is a common primary brain tumor with aggressive, lethal, and malignant characteristics. Its high proliferative and invasive nature leads to T- cell immunosuppression and drug inefficiency and hinders surgical resection. The OCT( right) shows the lipid-rich plaque (lip) with a low signal appearance and poorly delineated borders There is a need to compared with the signal-rich appearance of the fibrous plaque material (fib). (Courtesy of Meissner OA, investigate GBM in vivo in Rieber J, Babaryka G, et al: Intravascular optical coherence tomography: comparison with histopathology in preclinical animal models, atherosclerotic peripheral artery specimens. J Vasc lnterv Radiol 17:343–349, 2006. © SCVIR.) at the macro and micro levels, that are also OCT is also proving valuable in the differentiation potentially translatable to the clinic. Combined between cancerous and normal tissues as it is intravital microscopy using sensitive to the disruption of normal tissue confocal fluorescence (CF) architecture. The picture to the left shows the and optical coherence image of a sarcoma, or muscle tumor, obtained tomography (OCT) can be using (OCT). In the picture, the tissue looks used for the purpose. The healthy and normal on the left. To the right, the potential applications include surgical guidance structure appears cancerous and irregular. Images and monitoring of tumor like this one, obtained by using OCT in real-time, response to photodynamic can help detecting tumors early during image- therapy (PDT). Using this guided procedures. Due to its inherent imaging technique enables compatibility with the use of compact fiber-based real-time microvasculature probes, and the ability to construct portable imaging of brain tumors and normal brain tissue in order systems, OCT appears to be promising in the early detection of several types of cancer to track the tumor growth in clinics. OCT aided with computerized recognition and classification of tissue structure pattern in vivo and to has capabilities for in-vivo detection of bladder cancers, colon cancers, oral cancers monitor and quantify the and skin cancer. In detecting breast cancer for instance, compact optical fiber probes tissue responses to PDT permit access within the ductal structure of the breast or to a suspicious lesion via the treatment. A computerized tip of biopsy needle making possible to perform localized optical imaging of tissues at system based on tumor boundaries recognition the needle tip, this along with real-time feedback, has the potential to enhance the would provide a real-time guidance accuracy of the biopsy and reduce “miss-rate” compared to large-core needle and potentially available biopsy obtained under ultrasound guidance. 24-7 monitoring and Higher resolution and acquisition rates of OCT images improve real-time imaging quantification. capability. Given the data acquisition rates possible with the state-of-the-art OCT systems, rigorous human interpretation of every image is not possible in real-time. Thus high speed computer algorithms are essential for real-time feedback during biopsy or surgical guidance procedures to enable synchronization of motor rotation with the high-speed OCT frame acquisition in mechanically actuated probes, and to enable real-time tissue classification and suspicious object recognition.
- 6. 6 Increased speed and enhanced safety of visual based robot control An Example of a vision-based control task is for a robot arm to Visual Feedback control of robots acquire an unoriented object from a pallet without prior knowledge of the object position. A CCD camera attached to the Both industrial robot arms and mobile robots require sensing capability to robot arm provides visual adapt to new tasks without explicit intervention or reprogramming. Visual sensing capability. Images sensing capability of robots overcomes many of the difficulties of uncertain acquired by the camera are models and unknown environments which limit the domain of application of processed by a computer vision robots used without external sensory feedback. system in order to identify the The image-based structure is an approach to visual servo control, which uses image object and infer relationships features (e.g., image areas, and centroids) as feedback control signals, thus between the spatial position of eliminating a complex interpretation step (i.e., interpretation of image features to the object and the camera derive world-space coordinates). position. Such relative position information is used to guide the robot to acquire the object. The same problem arises in the navigation of a mobile robot with respect to objects in an unstructured environment using visual feedback. The Use of computer vision to infer position and orientation of objects or interpret general three- dimentional relationships in a In this approach, a 2-D image from a video sequence is processed in order to scene is a complex task extract image features. The extracted features are matched to precompiled object requiring extensive computing model features in order to identify, pick or track desired objects or provide an resources which may render absolute measure of the robot state (localization). In particular, the information gathered by a vision system about the environment makes it possible to detect robots slow in reacting to natural landmarks, navigate among unknown obstacles, and achieve a reactive unexpected events or robot behavior. Vision-based sensing however has some drawbacks, such as the constraints the robotic system need to recognize and extract a huge number of characteristic features from the with respect to speed. Using the image, an increased computational burden, and a critical dependence on lightning conditions of the environment. high speed parallel processing Concurrent Vision ApS delivers solutions in digital logic that compensate for approach, speed limits can be these drawbacks using a proven algorithm for feature extraction which is invariant removed and system safety can to scale, illumination, occlusion and viewing angle. Moreover, the solution from be enhanced. This can especially Concurrent Vision ApS implements a novel method of feature matching that accelerates the process up to real time speeds. be desired in robotic systems designed to remove defect objects from high speed automated production lines.
- 7. 7 Diverse Object Recognition Applications Object Recognition aid to the Blind Visio-haptic wearable systems for the blind Obstacle Avoidance: Object recognition and tracking algorithms for visio-haptic Object Recognition aids provide advance warning of information analysis, i.e., the conversion of visual data obstacles and allow the blind into haptic (tangible) features, can be utilized in wearable to find a safe, clear path. A assistive devices for blind individuals. Touch is an popular system is the important modality for individuals who are blind, but it is SonicGuide Also known as the Binaural Sensory Aid. This limited to the extent of one's reach. By estimating how an device uses ultrasound to object feels from its visual image, we are able to scan the space in front of the overcome this limitation. user and creates a stereo Common haptic devices and systems allow blind people audio signal that varies in pitch to indicate the distance recognize three-dimensional (3D) objects that exist in of obstacles. The system fits virtual environment. Such systems allow blind people to conveniently in the frame of a touch, grasp and manipulate objects that exist in the hap- pair of glasses. Although the tic enabled virtual environment. In the regard object rich information provided by the SonicGuide can be recognition reduces the overall time needed to understand the shape of objects and extremely useful, learning provide better immersion to the virtual environment. how to decode this signal requires significant effort There is also fear that the audio signal could mask Tracking Objects in Augmented Reality important environmental sounds. While audition is Augmented reality (AR) is a term for a live direct or indirect view of a physical already tapped for echolocation, the sense of real-world environment whose elements are merged with virtual computer- touch is largely unused while generated imagery - creating a mixed reality. The augmentation is conventionally traveling. It is thus possible interactive in real-time and in semantic to stimulate the skin without interfering with the normal context with environmental elements, activities and environmental such as sports scores on TV during a cues used by the blind. Moreover, it may be easier to match. With the help of advanced AR represent spatial information technology the information about the on the skin rather than through audition. surrounding real world of the user The best known such aids can becomes interactive and digitally usable. be classified as vision substitution systems since Artificial information about the they provide sufficiently rich environment and the objects in it can be information to be used. These stored and retrieved as an information systems use vision by CCD to scan the space and can be layer on top of the real world view. integrated in an earpiece on a The ability to track visible objects in real-time provides an invaluable tool for the hearing aid device as well in a haptic device. Such systems implementation of Augmented Reality. Once an object has been detected, it’s are required to be small, location in future frames can be used to position virtual content, and thus annotate lightweight and ultra low the environment. Object recognition and tracking solutions provided by Concurrent power consuming which can Vision ApS can effectively be used in real-time AR systems. be achieved by implementing algorithms in HW using the ASIC technology. Concurrent Vision ApS Steen Koldsø Moatasem Chehaiber steen.koldsoe@gmail.com moatasem@abs2real.dk . Mobile: + 4550168167 Mobile: +4530328964
- 8. Concurrent Vision ApS Hvidovrevej 44 2610 Rødovre +4530328964 moatasem@abs2real.dk Moatasem Chehaiber Company Details moatasem@abs2real.dk Steen Koldsø Moatasem Chehaiber: Senior Electronics Engineer and CTO steen.koldsoe@gmail.com Steen Koldsø: Senior Management Consultant and CEO MD. Mohammad Chehaiber: Consultant In Endocrinology MD. Tahseen Chouheiber: Consultant In Orthopedics MD. Mohammad Elhashimy: Consultant In General medicine Join the biomedical Engineering group on LinkeIn. A forum connecting biomedical engineers and biomedical imaging experts where people share biomedical engineering ideas based on biomedical image analysis. http://www.linkedin.com/groups?gid=2788350&trk=myg_ugrp_ovr