Television – Panoramic
Reexamination Certificate
1995-01-17
2001-06-05
Lee, Michael H. (Department: 2614)
Television
Panoramic
C348S207110, C348S206000, C348S147000, C348S335000, C382S293000
Reexamination Certificate
active
06243131
ABSTRACT:
TECHNICAL FIELD
This invention relates to the field of imaging systems generally and, more particularly, to the field of surveillance systems. Even more specifically, the present invention relates to a system and method for capturing the individual picture elements of an image, scanning the captured image, transforming at least one selected portion of the image to eliminate a predetermined distortion while magnifying, as appropriate, the selected portion, and outputing the image, either in analog or digital format and without the need for memory buffers. Preferably, the system is implemented in the form of a single integrated circuit incorporating an image sensor array and an application specific integrated circuit coupled thereto.
BACKGROUND ART
Video imaging is touching our lives in increasingly greater ways. For example, video imaging is used for security purposes, tele-conferencing, entertainment, robotics and endoscopic surgery, just to name a few. Video devices monitor and communicate to the masses.
One method of capturing a video image is through a fisheye lens. In such a method, three basic steps are required to yield an intelligible image. Namely, an image is captured using a video camera incorporating the fisheye lens, the image is processed in real-time to remove the optical distortion, and the image, or a portion of interest, is then reconstructed. A method and apparatus for electronically removing distortion from an image captured using a fisheye, wide angle or other lens for receiving a partially spherical image has been described in U.S. Pat. No. 5,185,667 and its progeny including U.S. Pat. Nos. 5,359,363 and 5,313,306 and U.S. application Ser. No. 08/189,585 filed Jan. 31, 1994 and Ser. No. 08/339,663, filed Nov. 14, 1994 all incorporated by reference as to any subject matter contained therein. Using the approach taught therein, the location of the desired output picture element in the input memory buffer is calculated and the input memory is sequenced through as directed by the distortion correcting calculations. One limitation of this approach is that it requires large quantities of fast memory to capture and store the incoming image in sequence so that the calculated addresses can rearrange the image removing the distortion. This method is required due to the standard scan approach used by the vast majority of commercial cameras.
Further, for higher resolution scan conversions, proportionately higher processing frequencies are required, for example, for capturing and digitizing the image, as well as increased capacity of high speed memory. These higher frequency components may be more expensive to implement and operate.
It is well known that conventional digital image arrays are primarily based on two technologies: Charged Coupled Diode (CCD) and Charge Injection Diode (CID) devices. Due to the standardized formats used in various countries—for example, N.T.S.C. is used in the United States and Japan, PAL is used in Europe, S.E.C.A.M. is used in Eastern Europe, the so-called “Grand Alliance” format and European 1250/50 format are suggested presently for high definition television in the United States and Europe respectively and the M.U.S.E. standard presently existent in Japan—for video transmission and reception, the scan sequencing on these arrays is typically top left to bottom right, sequencing through each pixel by row left to right and then by column. In conventional cameras, the captured image impinges on the imaging array in an undistorted manner, the scan thus providing an undistorted image.
For example, with the application of a fisheye lens to the camera, a distorted image impinges on the imaging array and a standardized scan results in a distorted image. The captured image is converted from analog to digital data and stored in a memory buffer. As taught in the Zimmerman '667 patent, the picture elements, or pixels, are scanned in an appropriate nonlinear sequence determined to yield a substantially non-distorted image. The particular pixels and the order in which they are scanned is dependant upon several factors including, for example, the orientation angle (zenith and azimuth angles from the direction of the lens), magnification, and rotation of the image. The transformed image is then converted back from digital to analog and stored in a second memory buffer. The analog data is output, for example, for recording, for display on a video display device or for compression and transmission to a remote location or otherwise utilized.
Charge Injection Diode (CID) imaging systems have been commercially available since the 1970's. CID imaging systems were introduced just prior to the introduction of Charged Coupled Diode (CCD) cameras. Of these competing technologies, the CCD technique has been the more widely used due to its ability to be mass produced and its simplified scanning control method. However, CID's remain in use in special applications due to their radiation tolerance and their ability to integrate light over long periods of time as a result of low light level sensitivity in the visible and infrared spectrums.
It is known that one unique property of CID's relative to CCD's is that they can be directly addressed on a pixel by pixel basis. However, directly addressing each pixel is costly due to the required circuitry needed to generate each address. Other problems that arise in the use of CID technology include interfacing transform hardware to a CID system at the sensor level and increasing the resolution of the CID to support increases in the zooming capability of the hardware.
Recently, new image sensor technology has emerged and has been described, for example, by Eric R. Fossum in his article, “Ultra Low Power Imaging Systems Using CMOS Image Sensor Technology” published in
Proceedings of the S.P.I.E.,
vol. 2267, Advanced Microdevices and Space Science Sensors (1994), incorporated by reference as to any subject matter deemed essential. Therein, a device herein referred to as an active pixel image sensor (APS) is described for manufacture according to complementary metal-oxide-semiconductor (CMOS) fabrication techniques. The APS is integratable with application specific integrated circuits on the same integrated circuit greatly improving access time. The APS image sensor may be provided with on-chip timing, control, signal chain and analog-to-digital conversion (ADC). Of particular importance is that the present device is active, i.e. not a passive device, and with built-in amplification, has improved noise immunity and light pick-up capability. Because of the movement in CMOS technology toward larger and larger capacity random access memories such as dynamic random access memories to 4, 8, 16 and soon 256 megabit capacity, it is entirely conceivable that access time and, just as importantly, resolution are greatly enhanced over either CID or CCD technologies as these devices are fashioned into arrays of greater and greater dimensions.
According to U.S. Pat. No. 5,200,818, issued Apr. 6, 1993, and incorporated herein by reference as to its entire contents, there is disclosed a partially spherical array (
FIG. 2
) or a circular array (
FIG. 3
) of CCD sensors which capture an image. It is believed that the captured image will exhibit little inherent distortion since each sensor of each array will capture a small portion of the overall image. The captured image represents the sum of the non-overlapping images captured by the several sensors. In other words, there would be little inherent distortion compared with that of an image captured utilizing a pinhole, fisheye, wide angle lens or other lens which introduces a predetermined distortion. The obvious disadvantage of the '818 patent is the present limited capability to manufacture such an array at reasonable cost.
Therefore, it is an object of this invention to provide a means for directly addressing each picture element of an analog image captured with an imaging device having a field of view, the picture elements being addressed in a non-linear sequen
Banner & Witcoff , Ltd.
Interactive Pictures Corporation
Lee Michael H.
LandOfFree
Method for directly scanning a rectilinear imaging element... does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for directly scanning a rectilinear imaging element..., we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for directly scanning a rectilinear imaging element... will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-2535022