Muhammad Saleem

Logo detection and recognition in video stream

Logos sometimes also known as trademark have high importance in todays marketing world. Products, companies and different gaming leagues are often recognized by their respective logos. In this paper, an implementation system for logo detection and recognition in video stream is presented. Feature set using SIFT (Scale Invariant Feature Transform) key-points is created for each query logo and the video frame being processed. The system uses a concept of bidirectional or two-way matching of features. Model fitting is carried out by using Random Sample Consensus (RANSAC) algorithm to suppress any outliers from two-way matching process. Also novel threshold scheme is used to finally make decision regarding presence of logo. Finally boundary is detected around the matched logo. The system provides various kinds of statistics like frequency of recognized logos, their visibility time and their locations in the video. Experiments were performed on various sports videos. The results of the experiments show that the developed system has achieved the precision rate of 89% and a recall of 99%.

Comparative analysis of invariant schemes for logo classification

Logo or Trademark is of high importance because it carries the goodwill of the company and the product. Products are mostly recognized by their brand logos. Their recognition is a major problem. A number of techniques are there for logo recognition. In this paper, a number of invariant techniques are compared to find out their effectiveness on various categories of brand logos. Techniques which were investigated are Hus Invariant Moments, Log-Polar Transform (LPT), Fourier-Mellin transform (FMT), Gradient Location-Orientation Histogram (GLOH). Experiments were performed on University of Marry Land (UMD) database. Results are given, along with recognition rate and time taken. Results show that GLOH performs the best with approximately 97% recognition rate but need a more computational time while on the other extreme FMT technique performs poorest with recognition rate with average of approx. 85% but with least computational time compared to all other techniques.