Ihre Aufgaben:

As a leading provider of camera-monitor-systems (CMS) for the automotive industry, Motherson R&D needs a large array of tools to record, replay, and process video data. These tools allow us to improve and evaluate our systems to achieve the best image quality in the industry and object detection performance.

Topic #1: Development of a simple Image Signal Processing Unit

The goal of this thesis is to develop an Image Signal Processing Unit that can be integrated in our existing video recording environment. This will enable us to record raw data from the cameras without a CMS system connected, to see the image while recording and generate the control commands for the cameras.

Your tasks:

• Investigate available options for real-time Image Signal Processing algorithms that can convert RAW video data from Bayer format
• Select a fitting algorithm and implement it using Intel Media SDK or other libraries, including operations in Bayer & RGB space such as White Balance Black Point Adjustment, White Balance, Lens Geometric Distortion etc.
• Generate feedback data to send to the camera (Exposure and Auto-White-Balance Data) to handle various illumination conditions based on attached metadata
• Integrate the algorithm in SMRs video tooling framework

Topic #2: Implementation of a real-time lossless compression/decompression algorithm for video data in CSI/RAW format

The goal of this thesis is to implement a tool for realtime lossless compression and decompression for video stream for our existing video recording tooling. This will allow us to achieve the same performance, but using less space when storing the data.

Your tasks:

• Investigate available options for real time lossless compression/decompression algorithm for CSI/RAW data
• Select a suitable algorithm, and implement it, using either OpenCL, AVX Intrinsics or CUDA.
• Integrate the algorithm in SMR’s framework

Topic #3: Implement a low latency video streaming application protocol over the internet

The goal of this thesis is to implement a tool for streaming video streams over the internet with low latency. This will help the distributed development team to develop image quality-related topics remotely.

Your tasks:

• Investigate available options for low latency streaming application protocols
• Select a suitable application, and implement it using WebTransport
• Integrate the algorithm in SMR’s framework

Anforderungen:

• Student of computer science, electrical engineering, microelectronics or mathematics or similar (#1/2/3)
• You have first experience in programming, preferably with C/C++ (#1/2)
• Ideally, you have first experience with Video Processing (#1)
• Ideally, you have experience with OpenGL, OpenCL, AVX or CUDA (#2)
• You have first experience in programming, preferably with JavaScript (#3)
• You have first experience with Networking (#3)
• Ideally, you have first experience with OpenGL (#3)