Different dress, same person
Problem being addressed
Person re-identification, the process of matching pedestrian images across different camera views, is an important task in visual surveillance. However, the process has its limitations when tracking a person at different places and at different time if that person (e.g., a criminal suspect) changes his/her clothes, causing most existing methods to fail, since they are heavily relying on color appearance and thus they are inclined to match a person to another person wearing similar clothes.
The researchers assume that a person wears clothes of similar thickness and, thus the shape of a person would not change significantly when the weather does not change substantially within a short period of time, and suggest new deep contour-sketch-based network for overcoming person re-identification under moderate clothing changes. Specifically, to quantify the contour sketch images effectively, the researchers introduce an algorithms to select the relatively invariant and discriminative contour patterns. They do not only quantitatively analyze the challenge of the clothing change problem by varying the degree of clothing change but also analyze the performance of person re-identification when clothing changes are combined with other challenges.
Advantages of this solution
The proposed method achieves the best accuracy among the compared methods including hand-crafted features and deep-learning-based methods for person re-identification under moderate clothing change.
Possible New Application of the Work
The suggested model can be successfully used in hospitals to keep track of the patients that may have memory issues or suffer from mental disorders. In case of lack of personnel the visual system that can recognise body contours can be extremely efficient in case the patients change their robes and try to move around the hospital.
Source URL: #############