Frame Marked with Production Information(FMPI), POIM(Program Oriented Information Frame) is an extension of FMPI in TV broadcast streams. usage:

FMPI.exe imagename imagetype  [Zip]

The input image format can be JPG(ImageType=1), BMP(ImageType=2), TIF(ImageType=3) and PGM(ImageType=4). The filename of output feature is imagename.fea. The current version only support color images. The totol dimension of feature is 141, including color, texture, and edge features. The code can only be used for purpose of research, however please acknowledge its use with a citation PCM06 or MM06. If you feel interest for the sourcecode, email me jqwang@nlpr.ia.ac.cn.

(Submitted to ICME09P)  [Experimental Results]

The problem of image summarization is to determine a smaller representation but faithfully represent the original visual image appearance. In this paper, we propose a context saliency based image summarization approach in a supervised manner. Since merely visual saliency as importance measure is not enough, we incorporate redundancy-based contrast analysis and geometric segmentation into context saliency through naive Bayesian inference. Then we introduce a grid-based piecewise linear image warping scaleplate to maintain the proportion of salient objects. We argue that the image summaries should be appraised with target device specification under perception constrains, and we adopt the sweet spot evaluation to generate a flexible model that automatically combines cropping and warping methods. Additionally, we explore potential extension on multiple applications such as video retargeting, digital matting, image browsing etc. Experimental results show comparable performance compared to the state-of-art on common data sets.

Pedestrian counting has been a challenging topic especially in video surveillance for a long time due to the view variations, scale changes and spatial occlusions. We formulate the problem of pedestrian counting as a joint maximum a posteriori (MAP) problem. Markov Chain Monte Carlo (MCMC) is utilized to search for an optimal configuration set to match the spatio-temporal context.

[TCSVT]Jinqiao Wang, Wei Fu, Jingjing Liu, and Hanqing Lu: Spatio-temporal Group Context for Pedestrian Counting.

[MMM'11]Jingjing Liu, Jinqiao Wang, and Hanqing Lu:   

Adaptive Model for Robust Pedestrian Counting.

In this project, we used the directional chamfer matching method for pedestrian counting, based on ceiling cameras. Circles of multiple sizes are used as the shape models of pedestrians, since seldom occlusions happen in such perspective. We tracked individuals and the trajectories in the ROI are counted as the number of pedestrians.