TY - GEN
T1 - Moving target location by neighbour-search algorithm
AU - Shen, Dinggang
AU - Wu, Yen
AU - Qi, Feihu
N1 - Funding Information:
Detection of moving target i n the first step of the proposed algorithm. First, we assume the environment of our task. In our task, the camera is allowed to move with any direction. The number of targets is permitted to be more than 1. Supposed that the movement of the targets on the image plane is small during the period between the two consecutive images. It is a reasonable assumption in most cases, because the distance between the targets and camera is usually much far. Thus the rough image areas of the moving targets,w.lth noise, Supported by the Climbing Programme-National Key Project for Foundamentsl Research in China, Grant NSC 92097.
Publisher Copyright:
© 1994 IEEE.
PY - 1994
Y1 - 1994
N2 - An effective moving object location algorithm using the neighbour-search algorithm (NSA) is proposed. The NSA can be used for real time applications. In order to locate the moving targets accurately, mapping of the target's statistics is suggested followed by the NSA. In fact, the NSA has the ability to locate any size targets effectively on the image plane. Owing to a large quantity of noise existing in our task, a fast median noise filter is suggested to remove it. Some noise may exist after filtering, and it is fortunate that the NSA can negate it. For a more accurate location, the NSA can be used twice for every target. In the first time the target is searched roughly and in the second time the NSA is used to search for the target accurately in the detected area by reducing the mapping ratio m. Experimental results show the effectiveness of our algorithm.
AB - An effective moving object location algorithm using the neighbour-search algorithm (NSA) is proposed. The NSA can be used for real time applications. In order to locate the moving targets accurately, mapping of the target's statistics is suggested followed by the NSA. In fact, the NSA has the ability to locate any size targets effectively on the image plane. Owing to a large quantity of noise existing in our task, a fast median noise filter is suggested to remove it. Some noise may exist after filtering, and it is fortunate that the NSA can negate it. For a more accurate location, the NSA can be used twice for every target. In the first time the target is searched roughly and in the second time the NSA is used to search for the target accurately in the detected area by reducing the mapping ratio m. Experimental results show the effectiveness of our algorithm.
UR - http://www.scopus.com/inward/record.url?scp=85064717565&partnerID=8YFLogxK
U2 - 10.1109/SIPNN.1994.344916
DO - 10.1109/SIPNN.1994.344916
M3 - Conference contribution
AN - SCOPUS:85064717565
T3 - ISSIPNN 1994 - 1994 International Symposium on Speech, Image Processing and Neural Networks, Proceedings
SP - 264
EP - 267
BT - ISSIPNN 1994 - 1994 International Symposium on Speech, Image Processing and Neural Networks, Proceedings
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 1994 International Symposium on Speech, Image Processing and Neural Networks, ISSIPNN 1994
Y2 - 13 April 1994 through 16 April 1994
ER -