Synthetic discriminant function automatic target recognition system augmented by LADAR
원문보기
IPC분류정보
국가/구분
United States(US) Patent
등록
국제특허분류(IPC7판)
F42B-015/01
F41G-007/00
출원번호
US-0250914
(1999-02-16)
발명자
/ 주소
Sims S. Richard F.
Pittman William C.
출원인 / 주소
The United States of America as represented by the Secretary of the Army
대리인 / 주소
Tischer
인용정보
피인용 횟수 :
35인용 특허 :
9
초록▼
Synthetic discriminant function automatic target recognition system augmented by LADAR combines, through the synthetic discriminant function (SDF), the active LADAR data of the potential target object with the passive infrared imagery of the target and background. In doing so, the system not only re
Synthetic discriminant function automatic target recognition system augmented by LADAR combines, through the synthetic discriminant function (SDF), the active LADAR data of the potential target object with the passive infrared imagery of the target and background. In doing so, the system not only recognizes and classifies the target but also provides the range profile of the target object by analyzing the amplitude of the reflected return signal when appropriate. The live target scene imagery in passive infrared is detected, filtered and subsequently complex multiplied with pre-existing synthetic discriminant function to produce a two-dimensional cross-correlated surface. Analogous process is performed on the active LADAR range and intensity images of the live target scene with corresponding pre-existing synthetic discriminant function for the same target pose and scale as in the passive infrared correlation step. Ultimately, the target information thus acquired in passive infrared and active LADAR is fused to obtain a more accurate determination of the target location and classification than is possible using either infrared or LADAR alone.
대표청구항▼
[ We claim:] [4.] An automatic target recognition system for detecting and tracking a target in a scenery, said system being resident in a seeker object and utilizing synthetic discriminant function to combine passive and active target imageries to determine accurate target location in the scenery a
[ We claim:] [4.] An automatic target recognition system for detecting and tracking a target in a scenery, said system being resident in a seeker object and utilizing synthetic discriminant function to combine passive and active target imageries to determine accurate target location in the scenery and enable precise tracking of the target, said system comprising: a tracker for tracking a detected target; a first sensor for collecting infrared image data of the target; a first pre-processor coupled to said first sensor to receive therefrom said infrared image data and perform convolution on said infrared image data to produce an infrared image filter; a first correlator coupled to said first pre-processor to receive therefrom said infrared image filter; a first synthetic discriminant function (SDF) memory module, said first SDF memory module being coupled to said first correlator and containing a multiplicity of pre-formed infrared synthetic discriminant function filters that are descriptive of various targets likely to be found in the scenery, each pre-formed infrared SDF filter describing a target from a given perspective and scale, a pre-selected infrared SDF filter from said first SDF memory module being input to said first correlator wherein said pre-selected infrared SDF filter is correlated with said infrared image filter of the live potential target to produce a first correlation surface; a second sensor for collecting LADAR range data of the target; a second pre-processor coupled to said second sensor to receive therefrom said LADAR range data and perform convolution on said LADAR range data to produce a LADAR range data filter; a second correlator coupled to said second pre-processor to receive therefrom said LADAR range data filter; a second synthetic discriminant function (SDF) memory module, said second SDF memory module being coupled to said second correlator and holding a multiplicity of pre-formed LADAR range synthetic discriminant function filters that are descriptive of various targets likely to be found in the scenery, each pre-formed LADAR range SDF filter describing a target from a given perspective and scale, a pre-selected LADAR range SDF filter from said second SDF memory module being input to said second correlator wherein said pre-selected LADAR range SDF filter is correlated with said LADAR range filter of the live potential target to produce a second correlation surface; a third sensor for collecting LADAR intensity data of the target; a third pre-processor coupled to said third sensor to receive therefrom said LADAR intensity data and perform convolution on said LADAR intensity data to produce a LADAR intensity data filter; a third correlator coupled to said third pre-processor to receive therefrom said LADAR intensity data filter; a third synthetic discriminant function (SDF) memory module, said third SDF memory module being coupled to said third correlator and holding a multiplicity of pre-formed LADAR intensity synthetic discriminant function filters that are descriptive of various targets likely to be found in the scenery, each pre-formed LADAR intensity SDF filter describing a target from a given perspective and scale, a pre-selected LADAR intensity SDF filter from said third SDF memory module being input to said third correlator wherein said pre-selected LADAR intensity SDF filter is correlated with said LADAR intensity filter of the live potential target to produce a third correlation surface; a classifier coupled to said tracker, said classifier being capable of classifying any target detected in the scenery; a means for fusing, said fusing means being simultaneously coupled to said first, second and third correlators to receive therefrom said first, second and third correlation surfaces, respectively, and being adapted for fusing said correlation surfaces to produce a composite correlation surface; a normalizer, said normalizer being coupled simultaneously to said first sensor, second sensor, third sensor and said fusing means, said normalizer receiving said infrared image data from said first sensor, said LADAR range data from said second sensor, said LADAR intensity data from said third sensor and said composite correlation surface from said fusing means and normalizing said composite correlation surface to produce therefrom a detection area; a statistics generator coupled between said normalizer and said classifier, said generator receiving said detection area from said normalizer and computing the mean and standard deviation of said area and inputing said mean and standard deviation to said classifier, said classifier, in response, producing a correlated target image and determining the target location and classification, thereby enabling said tracker to track a classified target with greater accuracy.
연구과제 타임라인
LOADING...
LOADING...
LOADING...
LOADING...
LOADING...
이 특허에 인용된 특허 (9)
McCoy Richard D. ; Stirman Charles E. ; O'Sullivan James N. ; Gordon Yvette M., Apparatus for and method of classifying patterns.
Corwin Thomas L. (McLean VA) Richardson Henry R. (Alexandria VA) Kuo Stanley D. (Arlington VA) Stefanick Tom A. (Arlington VA) Keeler R. Norris (McLean VA) Pflibsen Kent (Tucson AZ) Calmes Lonnie K. , Automatic target detection process.
Reis James J. (La Palma CA) Luk Anthony L. (Rancho Palos Verdes CA) Lucero Antonio B. (Anaheim CA) Garber David D. (Cypress CA), Target acquisition and tracking system.
Jenkins Gary Kim (Arlington TX) Evans Bruno Jack (Euless TX) Williams ; Jr. David Collis (Burleson TX) Bornowski Arthur Steven (Garland TX), Visual recognition system for LADAR sensors.
Jiang, Shyh-Biau; Chien, Hung-Chuan; Lee, Dong-Liang; Yang, Chi-Ming; Chang, Hsin-Ming; Huang, Chuan-Fu, Measuring system and method for detecting object distance by transmitted media with different wave velocities.
Haanpaa, Douglas; Cohen, Charles J.; Beach, Glenn J.; Jacobus, Charles J., Orientation invariant object identification using model-based image processing.
Haanpaa, Douglas; Cohen, Charles J.; Beach, Glenn J.; Jacobus, Charles J., Orientation invariant object identification using model-based image processing.
Haanpaa, Douglas; Cohen, Charles J.; Beach, Glenn J.; Jacobus, Charles J., Orientation invariant object identification using model-based image processing.
Haanpaa, Douglas; Cohen, Charles J.; Beach, Glenn J.; Jacobus, Charles J., Orientation invariant object identification using model-based image processing.
Stanfill, Steven Robert; Rutherford, Daniel B.; Beydoun, Hassan N.; Muise, Robert R.; Mahalanobis, Abhijit; Bhagavatula, Ramamurthy, System and method for passive automatic target recognition (ATR).
※ AI-Helper는 부적절한 답변을 할 수 있습니다.