This paper presents a new method for decision making based on the fusion of audio-visual evidence... more This paper presents a new method for decision making based on the fusion of audio-visual evidences. Evidences fusion is characterized that the decision will be more accurate and specific because it does not depend on one evidence's alone as in the probabilistic approach. The decision-making process depends on decisional separation of conflict conditionally between contributions of several independent sources of information represented by audio and images. In order to provide effective of evidence fusion, one must employ an analytical framework that can capture the uncertainty inherent in audio and visual data. In particular, feature extraction of audio and visual data results in propositions that inherently possess significant semantic ambiguity. An evidence fusion must be able to exploit the respective advantages of audio and visual data while mitigating their particular weaknesses.
In this paper, an intelligent advisory system is used to make the decision depend on extracting t... more In this paper, an intelligent advisory system is used to make the decision depend on extracting the features of events by using artificial intelligence techniques. The proposed system was applied in the training in medical institutions in the field of physical therapy. X-ray images and their corresponding reports are used as events. For X-ray images, the Co-occurrence matrix is used to extract the features and Evidential K-Nearest Neighboring Rule has used to features classification then stored in the database. Whereas, the hidden Markov model is used to recognize the handwritten reports after their processing. The probability theory is used to solve the feature's overlapping problem that makes it easier to knowledge base formation by experts. The proposed system has been applied to different and several pathological patients of bone fractures and jaw and dental injuries. Comparing the results with the experts has shown the high accuracy that characterizes the proposed system.
This paper presents a new method for decision making based on the fusion of audio-visual evidence... more This paper presents a new method for decision making based on the fusion of audio-visual evidences. Evidences fusion is characterized that the decision will be more accurate and specific because it does not depend on one evidence's alone as in the probabilistic approach. The decision-making process depends on decisional separation of conflict conditionally between contributions of several independent sources of information represented by audio and images. In order to provide effective of evidence fusion, one must employ an analytical framework that can capture the uncertainty inherent in audio and visual data. In particular, feature extraction of audio and visual data results in propositions that inherently possess significant semantic ambiguity. An evidence fusion must be able to exploit the respective advantages of audio and visual data while mitigating their particular weaknesses.
In this paper, an intelligent advisory system is used to make the decision depend on extracting t... more In this paper, an intelligent advisory system is used to make the decision depend on extracting the features of events by using artificial intelligence techniques. The proposed system was applied in the training in medical institutions in the field of physical therapy. X-ray images and their corresponding reports are used as events. For X-ray images, the Co-occurrence matrix is used to extract the features and Evidential K-Nearest Neighboring Rule has used to features classification then stored in the database. Whereas, the hidden Markov model is used to recognize the handwritten reports after their processing. The probability theory is used to solve the feature's overlapping problem that makes it easier to knowledge base formation by experts. The proposed system has been applied to different and several pathological patients of bone fractures and jaw and dental injuries. Comparing the results with the experts has shown the high accuracy that characterizes the proposed system.
Uploads