Implementing Content Based Video Retrieval Using Speeded-Up Robust Features

Abstract

In this paper, we propose an implementation of Content-Based Video Retrieval (CBVR) using Speeded-Up Robust Features (SURF). Given an image as a query, the application looks through a set of videos and pick the ones contain frame similar to the image query. Our objective is to measure the performance of the algorithm. The performance is measured using three variables: recall, precision, and running time. We used two sets of samples to perform the test: in-frame and not-in-frame. Furthermore, we limit the samples only to contain these 5 categories: body parts, kitchen and eating utensils, fruits, and pets. The test shows the program gives a 57.75% average recall value and 37.5% precision value for not-in-frame test, while the in-frame-test gives 51% and 59% for recall and precision value respectively. Moreover, the running time data shows there is no relation between in-frame/not-in-frame and speed. Running time performance highly depends on the image query and the length of the video.

Efficient Content Based Video Retrieval System by Applying AlexNet on Key Frames

“…And show that has SSVH approach has suggestively outclassed the state of the art systems and achieve the presently best result on the task. Tarigan et al (2018) invented an implementation of CBVR using SURF. In query give an image, the system search videos in database and retrieve similar videos with query image that matched with video frames.…”

A Deep Analysis of Image Based Video Searching Techniques

“…In Not-in frame, average precision was 25%, average recall was 66% and running time performance data gave 73.56 second. For in-frame, precision value was 59%, recall value was 51% and running time was 121.67 second in future, research will be on understanding the performance of SURF which is based on some specifics kind of videos [8] A technique was implemented to solve top-k video retrieval problem through a query image. The basic purpose of this model is to retrieve best matched video from large video database using image as query.…”