There are many visual inspection and sensing applications where both a high resolution image and a depth-map of the
imaged object are desirable at high speed. Presently available methods to capture 3D data (stereo cameras and structured
illumination), are limited in speed, complexity, and transverse resolution. Additionally these techniques rely on a
separated baseline for triangulation, precluding use in confined spaces. Typically, off the shelf lenses are implemented
where performance in resolution, field-of-view, and depth of field are sacrificed in order to achieve a useful balance.
Here we present a novel lens system with high-resolution and wide field-of-view for rapid 3D image capture. The design
achieves this using a single lens with no moving parts. A depth-from-defocus algorithm is implemented to reconstruct
3D object point clouds and matched with a fused image to create a 3D rendered view.
We have developed a standoff iris biometrics system for improved usability in access-control applications. The system
employs an eye illuminator, which is composed of an array of encapsulated near-infrared light emitting diodes (NIRLEDs),
which are triggered at the camera frame rate for reduced motion blur and ambient light effects. Neither the
standards / recommendations for NIR laser and lamp safety, nor the LED-specific literature address all the specific
aspects of LED eye-safety measurement. Therefore, we established exposure limit criteria based on a worst-case scenario
combining the following: the CIE/ANSI standard/recommendations for exposure limits; concepts for maximum
irradiance level and for strobing from the laser safety standards; and ad-hoc rules minimizing irradiance on the fovea, for
handling LED arrays, and for LED mounting density. Although our system was determined as eye safe, future variants
may require higher exposure levels and lower safety margins. We therefore discuss system configuration for accurate
LED radiometric measurement that will ensure reliable eye-safety evaluation. The considerations and ad hoc rules
described in this paper are not, and should not be treated as safety recommendations.
This paper presents an overview of Intelligent Video work currently under development at the GE Global Research Center
and other research institutes. The image formation process is discussed in terms of illumination, methods for automatic
camera calibration and lessons learned from machine vision. A variety of approaches for person detection are presented.
Crowd segmentation methods enabling the tracking of individuals through dense environments such as retail and mass
transit sites are discussed. It is shown how signature generation based on gross appearance can be used to reacquire targets
as they leave and enter disjoint fields of view. Camera calibration information is used to further constrain the detection
of people and to synthesize a top-view, which fuses all camera views into a composite representation. It is shown how
site-wide tracking can be performed in this unified framework. Human faces are an important feature as both a biometric
identifier and as a method for determining the focus of attention via head pose estimation. It is shown how automatic pan-tilt-
zoom control; active shape/appearance models and super-resolution methods can be used to enhance the face capture
and analysis problem. A discussion of additional features that can be used for inferring intent is given. These include
body-part motion cues and physiological phenomena such as thermal images of the face.
A novel technique for the detection and enhancement of microcalcifications in digital tomosynthesis mammography (DTM) is presented. In this method, the DTM projection images are used directly, instead of using a 3D reconstruction. Calcification residual images are computed for each of the projection images. Calcification detection is then performed over 3D space, based on the values of the calcification residual images at projection points for each 3D point under test. The quantum, electronic, and tissue noise variance at each pixel in each of the calcification residuals is incorporated into the detection algorithm. The 3D calcification detection algorithm finds a minimum variance estimate of calcification attenuation present in 3D space based on the signal and variance of the calcification residual images at the corresponding points in the projection images. The method effectively detects calcifications in 3D in a way that both ameliorates the difficulties of joint tissue/microcalcification tomosynthetic reconstruction (streak artifacts, etc.) and exploits the well understood image properties of microcalcifications as they appear in 2D mammograms. In this method, 3D reconstruction and calcification detection and enhancement are effectively combined to create a calcification detection specific reconstruction. Motivation and details of the technique and statistical results for DTM data are provided.
A novel technique for imaging spectroscopy is introduced. The technique makes use of an optical imaging system with a segmented aperture and intensity detector array on the imaging plane. The point spread function (PSF) of such a system can be adjusted by modifying the path lengths from the subapertures to the image plane, and the shape of the resulting point spread function will vary as a function of wavenumber. An image reconstruction approach is taken to convert multiple recorded pan-chromatic images with different wavenumber-varying point spread functions into a hyperspectral data set. Thus, the technique described here is a new form of computed imaging.
Optical imaging systems are often limited in resolution, not by the imaging optics, but by the light intensity sensors on the image formation plane. When the sensor size is larger than the optical spot size, the effect is to smooth the image with a rectangular convolving kernel with one sample at each non-overlapping kernel position, resulting in aliasing. In some such imaging systems, there is the possibility of collecting multiple images of the same scene. The process of reconstructing a de-aliased high-resolution image from multiple images of this kind is referred to as “super-resolution image reconstruction.” We apply the POCS method to this problem in the frequency domain. Generally, frequency domain methods have been used when component images were related by subpixel shifts only, because rotations of a sampled image do not correspond to a simple operation in the frequency domain. This algorithm is structured to accommodate rotations of the source relative to the imaging device, which we believe helps in producing a well-conditioned image synthesis problem. A finely sampled test image is repeatedly resampled to align with each observed image. Once aligned, the test and observed images are readily related in the frequency domain and a projection operation is defined.
An important telemedicine application is the perusal of CT scans (digital format) from a central server housed in a healthcare enterprise across a bandwidth constrained network by radiologists situated at remote locations for medical diagnostic purposes. It is generally expected that a viewing station respond to an image request by displaying the image within 1-2 seconds. Owing to limited bandwidth, it may not be possible to deliver the complete image in such a short period of time with traditional techniques. In this paper, we investigate progressive image delivery solutions by using JPEG 2000. An estimate of the time taken in different network bandwidths is performed to compare their relative merits. We further make use of the fact that most medical images are 12-16 bits, but would ultimately be converted to an 8-bit image via windowing for display on the monitor. We propose a windowing progressive RoI technique to exploit this and investigate JPEG 2000 RoI based compression after applying a favorite or a default window setting on the original image. Subsequent requests for different RoIs and window settings would then be processed at the server. For the windowing progressive RoI mode, we report a 50% reduction in transmission time.
A method for registering terrain range images is described. A smooth surface is fit to range data to provide a continuous mapping to second order surface intrinsics. Level sets of such intrinsics are found as features. These features enable two accurate means of comparison: arclength and a circular correlation of the distance from the discrete points on the level set to its centroid. These two methods of feature matching allow a matching algorithm with computational complexity O(N log N). The solution of the transformation from the matches employs a weighted least squares technique and removes matching errors by cross validation. The method is tested on simulated range images and results are given.