"Target Tracking: Correlation filter One" come back, dig a hole first.

Objective

Since December 2015, retreat cultivation, concentrated on the target tracking field of the relevant filters, now finally can be a temporary end. Now I compared to last year's I, self-feeling although not a reborn, but indeed learned a lot of knowledge, in the field of machine vision is no longer a green and not the door of the birds, now the bird, has been able to stumble off the takeoff.

With this blog post as the dividing line, before the post in my view is really a pattern Tucson broken, you crossing big can ignore the previous blog post.

From today onwards, will try to spare time to seriously write a blog, the time of the research results and experience to share to everyone. Today, dig a hole, content may be updated very slowly, because immediately three of the people, after the summer will find work, has been busy with the basic knowledge of the ingestion and tracking algorithm research, and lazy cancer late, coupled with the second disease committed two months, fascinated by the DOTA2 (online game ruined life, DotA poor three generations), wasted a lot of valuable time , did not do any substantive work, there is no engineering experience, to find a job before the implementation of my research in the embedded platform, is to master a meal technology, time is very urgent.

Here is a brief description of my Learning environment:

In a mechanics and engineering thermal Physics Laboratory, originally engaged in control and electronic circuit of people, but I dictionary the choice of machine vision target tracking in this direction, the whole laboratory more than 100 people only I am a person to engage in this, the mentor to this is also smattering, to my training also near stocking, no project , no group, brother elder sister do not engage in this, no brother sister, I am a man alone, due to less communication, in the computer college know peers and not very familiar, because seldom meet, so very little communication

Correlation Filter

For the current rounds in the field of tracking, it is necessary to understand the method of the general development process, in this first recommended a few paper, interested students can go first to see.
1. Bolme D S, Beveridge J R, Draper B A, et al. Visual Object tracking using adaptive correlation filters[c]//computer Vis Ion and Pattern Recognition (CVPR), IEEE Conference on. IEEE, 2010:2544-2550.
2. Henriques J F, Caseiro R, Martins P, et al exploiting the circulant structure of tracking-by-detection with Kernels[m] Computer VISION–ECCV 2012. Springer Berlin Heidelberg, 2012:702-715.
3. Henriques J F, Caseiro R, Martins P, et al high-speed tracking with kernelized correlation filters[j]. Pattern analysis and Machine Intelligence, IEEE transactions on, 2015, 37 (3): 583-596.
4. Chen Z, Hong z, Tao D. An experimental Survey on Correlation filter-based tracking[j]. ARXIV preprint arxiv:1509.05520, 2015.

The next few posts, I will be in front of the above 3 articles as the basis, as far as possible to communicate with you the basic type of correlation filter and its subsequent improvement. Let's stop today and start moving bricks.

"Target Tracking: Correlation filter One" come back, dig a hole first.