Software defined camera for outdoor surveillance

Traditional cameras are not flexible and also unaware of their environment. They cannot provide constant and standard image quality in every environment because of their fixed (non-programmable) internal architecture. Along with image quality, image pre-processing is also crucial. The images need to

Project Title

Software defined camera for outdoor surveillance

Project Area of Specialization

Computer Science

Project Summary

Traditional cameras are not flexible and also unaware of their environment. They cannot provide constant and standard image quality in every environment because of their fixed (non-programmable) internal architecture. Along with image quality, image pre-processing is also crucial. The images need to be pre-processed before they can be used for useful results. For this purpose, FPGAs are used which are programmable semiconductor devices that are based around a matrix of Configurable Logic Blocks (CLBs) connected through programmable interconnects. As opposed to Application Specific Integrated Circuits (ASICs), where the device is custom built for the particular design, FPGAs can be programmed to the desired application or functionality requirements. Along with the FPGA, Omnivision OV5640 image sensor is used to 'preprocess' the camera input before the capture of images. After that, the images were examined in MATLAB, and specific parameters were determined. Brightness, color, saturation, sharpness, and luminance were among the criteria. The parameters were utilized to identify the most important ones, which could then be manipulated during pre-processing. The camera was put in a fixed position at one particular location and was programmed to generate a live feed and take photographs at regular intervals. Before using the FPGA board and a connected camera, camera equipped with sensor OV5640 was to be used for the purpose of environmental profiling which wasn't available, so after searching various resources, we found out that that the image sensor in Pi camera i.e. OV5647 was similar in many aspects. Sensor parameters can be modified using command line interface. Parameter values range between 0 to 100 where 0 represented minimum and 100 maximum value. This is proof of concept that sensor state is modifiable and not fixed like traditional and conventional cameras. The implemented solution provides a way to realize a design on the FPGA board to pre-process the live video stream being captured in real-time. The proposed design for this can be extended with further image processing blocks and environment adaptable capturing features for further applications as well.

Project Objectives

To study and determine exact parameters which can be manipulated to produce high quality images.

• To integrate different models of the environment for an efficient solution.

• To exploit the internal architecture of the image sensor.

• To develop image pre-processing IP cores and implement a solution that will integrate all the cores.

• To test the software-defined camera for outdoor surveillance applications under various environmental conditions.

Project Implementation Method

Our project’s aim is to make camera smart enough, such that it produce high quality images in varying environment. The methodology we will use to achieve our target comprise of:

• Environmental Profiling:

Conventional cameras will be used for capturing images in different environments at regular intervals. These images help in determining the most varying parameters.

• Image Sensor Integration:

PCAM 5C module will be connected to the FPGA board and a set of open-source Vivado IP cores will be used for its configuration in software.

• Image Processing Cores using Vivado HLS:

Vivado High Level Synthesis will be used for generation of different soft IP cores.

Benefits of the Project

This project provide ease to customers of any environment varying image processing application e.g.

• In surveillance systems, as it is used to recognize the suspected person or vehicle from a still image or video.
• In airports, automated systems could be used to check luggage.
• In biometric recognition, to identify an individual based on some biological feature such as eye, ear, fingerprint, etc.

Technical Details of Final Deliverable

A remotely deployable software-defined camera solution that can adjust to different conditions, producing high-quality output video stream with minimal processing time, thus lessening the load of image pre-processing on server-side applications.

Final Deliverable of the Project

HW/SW integrated system

Core Industry

Security

Other Industries

Core Technology

Others

Other Technologies

Sustainable Development Goals

Decent Work and Economic Growth, Industry, Innovation and Infrastructure, Partnerships to achieve the Goal

Required Resources

If you need this project, please contact me on contact@adikhanofficial.com