scispace - formally typeset
Search or ask a question
Author

Wenjun Zhang

Bio: Wenjun Zhang is an academic researcher from Shanghai University. The author has contributed to research in topics: Optical flow & Image tracing. The author has an hindex of 1, co-authored 3 publications receiving 3 citations.

Papers
More filters
Proceedings ArticleDOI
19 Jul 2013
TL;DR: This paper proposes a method that first to determine the like-face area in the video frame with local SMQT characteristics, then positioning the detected human face feature point with the modified ASM, and grouping feature points based on their characteristics, tracking them by using optical flow method, elastic graph matching, and binary respectively.
Abstract: In the basic study and analysis of the human face detection and feature point location and tracking algorithm in video sequence, this paper proposes a method that first to determine the like-face area in the video frame with local SMQT characteristics; then positioning the detected human face feature point with the modified ASM, which is improved by changing the 1D texture model which is easier to fall into minimum to 2D texture model; finally, grouping feature points based on their characteristics, tracking them by using optical flow method, elastic graph matching, and binary respectively. This method was tested to show good positioning of facial features based on fast detection, and gain well tracking results.

1 citations

Proceedings ArticleDOI
01 May 2012
TL;DR: A unified framework is proposed to detect the shot boundaries and extract the keyframe of a shot through independent component (IC) analysis feature space and presents a new metric, image complexity, to extract keyframe in a shot which is computed by ICs.
Abstract: In recent years, Music video data is increasing at an astonishing speed. Shot segmentation and keyframe extraction constitute a fundamental unit in organizing, indexing, retrieving video content. In this paper a unified framework is proposed to detect the shot boundaries and extract the keyframe of a shot. Music video is first segmented to shots by illumination-invariant chromaticity histogram in independent component (IC) analysis feature space .Then we presents a new metric, image complexity, to extract keyframe in a shot which is computed by ICs. Experimental results show the framework is effective and has a good performance.© (2012) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.

Cited by
More filters
Journal ArticleDOI
TL;DR: Wang et al. as discussed by the authors found that similar logos can cause lower rate of recognition, increase in the chance of consumers to misunderstand or mix up, and behavior of tort that lead to damage enterprise itself and customers.
Abstract: Under the global competition, using integrated marketing of brands to raise the economic effect has become an important edge. Not only being distinguishable from competitors, brands can also become a key point of stimulating the behavior of purchase. Therefore the design of trademark images is extraordinarily important. Similar logo can cause lower rate of recognition, increase in the chance of consumers to misunderstand or mix up, and behavior of tort that lead to damage enterprise itself and customers. So far, in the world-famous periodicals and documents, we have not found related research of logo similarity; therefore, this will be a groundbreaking research. In this research, we choose logo of Coca Cola which has kept in leading position under the evaluation of brand consultant company “Interbrand”. We produced twelve images which including the original logo that were done by mixing up Coca Cola logo. In addition, we asked forty graphic designers who have worked in related field for over two years to assist in building distribution map of logo so that we can figure the main reason of mix-up in trademark image up. Enterprises can use this result to draft a strategy that make themselves become more competitive and unique.

3 citations

Proceedings Article
01 Jan 2008
TL;DR: Through the improved ASM algorithm the local minima problem can be solved efficiently, the detailed local information of face feature points is extracted and localized the points accurately.
Abstract: Active Shape Models(ASM) is one of powerful tools for facial feature location and face recognition.However,the performance of ASM is often influenced by some factors such as the initial location,illumination and so on,which will frequently lead to the local minima in optimization.This paper proposes an improved ASM algorithm on base of traditional one.First,using Adaboost method initial positioning of the significant facial region in order to carry out follow-up search in the location of the binding.Second,1D texture model of the original ASM is extended to 2D texture model which is based on the kernel probability density estimation model;At last,we add fringe restraint in the model of partial grey level,so that the points with stronger fringe information could have more possibility to become the perfect candidate points.Through the improved method the local minima problem can be solved efficiently,extracted the detailed local information of face feature points and localized the points accurately.Experiments on a database containing 200 labeled face images show that the improved ASM is more accurate and robust than the conventional one.

1 citations