There is one method, called the block gradient method here, whereby to detect image movements from an imagesequence. In this method, a relation between image derivatives and a movement is used with an assumption that movements in a block have the same values. This method has an advantage over the computing cost, compared with other methods, such as the block matching method. However, it also involves a general problem, which is that the reliability of a detected movement varies with varying conditions. Therefore, it is important to obtain some indices indicating the reliability of the detected movements. The reason is that, without such information, it is impossible to make good use of the movements. In this paper, the reliability of a movement detected by the block gradient method is discussed and indices indicating it are proposed. One index indicates the support level for the assumption used in the block gradient method. Two other indices indicate the image pattern condition, or the aperture problem level, in other words. The detected movement accuracy is rather directly related to this condition. The indices effectiveness is shown by experiments. The indices are expected to offer helpful information to the sequential processes using the movements. In addition, a movement stabilization process is described. This process smooths a movement field in the image, using the detected movements and two of the three indices. It is useful, when the movement detection process is used independently.
The copyright of the original papers published on this site belongs to IEICE. Unauthorized use of the original or translated papers is prohibited. See IEICE Provisions on Copyright for details.
Copy
Naoya OHTA, "Image Movement Detection with Reliability Indices" in IEICE TRANSACTIONS on Information,
vol. E74-D, no. 10, pp. 3379-3388, October 1991, doi: .
Abstract: There is one method, called the block gradient method here, whereby to detect image movements from an imagesequence. In this method, a relation between image derivatives and a movement is used with an assumption that movements in a block have the same values. This method has an advantage over the computing cost, compared with other methods, such as the block matching method. However, it also involves a general problem, which is that the reliability of a detected movement varies with varying conditions. Therefore, it is important to obtain some indices indicating the reliability of the detected movements. The reason is that, without such information, it is impossible to make good use of the movements. In this paper, the reliability of a movement detected by the block gradient method is discussed and indices indicating it are proposed. One index indicates the support level for the assumption used in the block gradient method. Two other indices indicate the image pattern condition, or the aperture problem level, in other words. The detected movement accuracy is rather directly related to this condition. The indices effectiveness is shown by experiments. The indices are expected to offer helpful information to the sequential processes using the movements. In addition, a movement stabilization process is described. This process smooths a movement field in the image, using the detected movements and two of the three indices. It is useful, when the movement detection process is used independently.
URL: https://global.ieice.org/en_transactions/information/10.1587/e74-d_10_3379/_p
Copy
@ARTICLE{e74-d_10_3379,
author={Naoya OHTA, },
journal={IEICE TRANSACTIONS on Information},
title={Image Movement Detection with Reliability Indices},
year={1991},
volume={E74-D},
number={10},
pages={3379-3388},
abstract={There is one method, called the block gradient method here, whereby to detect image movements from an imagesequence. In this method, a relation between image derivatives and a movement is used with an assumption that movements in a block have the same values. This method has an advantage over the computing cost, compared with other methods, such as the block matching method. However, it also involves a general problem, which is that the reliability of a detected movement varies with varying conditions. Therefore, it is important to obtain some indices indicating the reliability of the detected movements. The reason is that, without such information, it is impossible to make good use of the movements. In this paper, the reliability of a movement detected by the block gradient method is discussed and indices indicating it are proposed. One index indicates the support level for the assumption used in the block gradient method. Two other indices indicate the image pattern condition, or the aperture problem level, in other words. The detected movement accuracy is rather directly related to this condition. The indices effectiveness is shown by experiments. The indices are expected to offer helpful information to the sequential processes using the movements. In addition, a movement stabilization process is described. This process smooths a movement field in the image, using the detected movements and two of the three indices. It is useful, when the movement detection process is used independently.},
keywords={},
doi={},
ISSN={},
month={October},}
Copy
TY - JOUR
TI - Image Movement Detection with Reliability Indices
T2 - IEICE TRANSACTIONS on Information
SP - 3379
EP - 3388
AU - Naoya OHTA
PY - 1991
DO -
JO - IEICE TRANSACTIONS on Information
SN -
VL - E74-D
IS - 10
JA - IEICE TRANSACTIONS on Information
Y1 - October 1991
AB - There is one method, called the block gradient method here, whereby to detect image movements from an imagesequence. In this method, a relation between image derivatives and a movement is used with an assumption that movements in a block have the same values. This method has an advantage over the computing cost, compared with other methods, such as the block matching method. However, it also involves a general problem, which is that the reliability of a detected movement varies with varying conditions. Therefore, it is important to obtain some indices indicating the reliability of the detected movements. The reason is that, without such information, it is impossible to make good use of the movements. In this paper, the reliability of a movement detected by the block gradient method is discussed and indices indicating it are proposed. One index indicates the support level for the assumption used in the block gradient method. Two other indices indicate the image pattern condition, or the aperture problem level, in other words. The detected movement accuracy is rather directly related to this condition. The indices effectiveness is shown by experiments. The indices are expected to offer helpful information to the sequential processes using the movements. In addition, a movement stabilization process is described. This process smooths a movement field in the image, using the detected movements and two of the three indices. It is useful, when the movement detection process is used independently.
ER -