Tuesday, 20 October 2015
Smart Photo: A Resource-Aware Crowd Sourcing Approach for Image Sensing With Smart Phones
Photos obtained via crowd sourcing can be used in many critical applications. Due to the limitations of communication bandwidth, storage and processing capability, it is a challenge to transfer the huge amount of crowd sourced photos. To address this problem, we propose a framework, called Smart Photo, to quantify the quality (utility) of crowd sourced photos based on the accessible geographical and geometrical information (called metadata) including the Smartphone’s orientation, position and all related parameters of the built-in camera. From the metadata, we can infer where and how the photo is taken, and then only transmit the most useful photos. Four optimization problems regarding the tradeoffs between photo utility and resource constraints, namely Max-Utility, online Max-Utility, Min-Selection and Min-Selection with k-coverage, are studied. Efficient algorithms are proposed and their performance bounds are theoretically proved. We have implemented Smart Photo in a test bed using Android based smart phones, and proposed techniques to improve the accuracy of the collected metadata by reducing sensor reading errors and solving object occlusion issues. Results based on real implementations and extensive simulations demonstrate the effectiveness of the proposed algorithms.
The aim of this paper is propose a framework, called Smart Photo, to quantify the quality (utility) of crowd sourced photos based on the accessible geographical and geometrical information (called metadata) including the Smartphone’s orientation, position and all related parameters of the built-in camera.
The scope of this paper is implemented Smart Photo in a test bed using Android based smart phones, and proposed techniques to improve the accuracy of the collected metadata by reducing sensor reading errors and solving object occlusion issues.
The major challenges faced by these applications are as follows. The first is how to characterize the quality (usefulness) of crowd sourced photos in a way that is both meaningful and resource friendly. Most content-based image processing techniques such as may demand too much computational and communication resources at both user and server end. On the other hand, existing solutions from description based techniques either categorize photos based on user defined tags, or prioritize them by the GPS location . Obviously, tagging each photo manually is not convenient and may discourage public participation. GPS location itself may not be sufficient to reveal the real point of interest. Even at the same location, smart phones facing different directions will have different views.
· Limitations of communication bandwidth, storage and processing capability.
· To identify the most relevant data and eliminate redundancy becomes an important issue.
In this project, propose Smart Photo, a novel framework to evaluate and optimize the selection of crowd sourced photos, based on the collected metadata from the smart phones. We formulate the Max-Utility problem for bandwidth constrained networks, and then extend it into an online optimization problem. We study the Min-Selection problem for redundancy reduction, and also extend it to the case where better coverage (e.g., k-coverage) is desired. Moreover, we propose efficient solutions, and find the performance bounds in terms of approximation or competitive ratios for the proposed algorithms. We have implemented Smart Photo in a test bed using Android based smart phones. We make use of multiple embedded sensors in off-the-shelf smart phones, and propose a series of methods to fuse data, correct errors, and filter out false information, to improve the accuracy of the collected metadata. Finally, the performance of the proposed algorithms is evaluated through real implementations and extensive simulations.
· Resource constraint of bandwidth, storage and processing capability limits the number of photos that can be uploaded to the server.
· Consider having certain level of redundancy in case better coverage is needed.
· Speed - 1.1 Ghz
· RAM - 256 MB(min)
· Hard Disk - 20 GB
· Floppy Drive - 1.44 MB
· Key Board - Standard Windows Keyboard
· Mouse - Two or Three Button Mouse
· Monitor - SVGA
· Operating System :Android OS
· Front End : JAVA
· Database : SqLite
· Tool :Eclipse
Wang, Y. , Hu, W. , Cao, G. Wu, Y. “Smart Photo: A Resource-Aware Crowd sourcing Approach for Image Sensing with Smart phones” IEEE Transactions on Mobile Computing, Volume PP , Issue 99 June 2015