Automotive Imaging

Camera sensor are used widely in automotive safety applications such as ADAS and in full autonomous cars for surround mapping. Camera sensors in cars are exposed to extreme condition in terms of motion blurs, noise, lighting and more. These things make it a necessity to have a dedicated algorithms to take care of all the shortcomings, helping the other depended algorithm to have a constant video feed irrespective of the outside lighting condition such as tunnel and indoor parking lots.

Need for dedicated Auto Camera Function in Automotive Imaging

Radar based object detection SDK

Automotive cameras undergo dynamic environment and scene types with extreme light and illuminant variations

This requires sensor exposure and ISP parameters to be adapted dynamically depending on light conditions

Radar based object detection SDK

Software based functions are crucial in estimating the sensor exposure (for well exposedness), managing dynamic range and estimating illuminant conditions automatically based image statistics driving the overall dynamic parameter settings for ISP blocks

2A Algorithm for Automotive Imaging

Auto Exposure
  • Automatic adjustments of the image brightness according to the amount of light that reaches the camera image sensor
  • Flexible in using statistics data from ISP or from sensor
  • Supports different metering modes including multi ROI based metering mode
  • Exposure control for correcting flicker under 50Hz/60Hz line source
  • Supports configuring multiple exposure times in case of WDR/HDR and HDR+LFM sensors
  • Scalable design to port on different embedded platforms
  • MISRA-C compliant
Auto white Balance
  • Auto white balance is the process of removing unrealistic color casts, so that objects which appear white in real world are rendered white in the captured image or video
  • Estimates color temperature of illuminant/scene
  • Flexible in using statistics data from ISP or from sensor
  • Supports both calibration based and non calibration based approach
  • Offline PC based AWB calibration tool
  • MISRA-C compliant

Test Results

2A and Image Quality Tuning on TDA4x & AR0233AT

Supported Platforms and Sensors

Platform's: TDA4x, SA81xx

Camera sensor’s: OV2312, OG01A

PathPartner has developed Auto Exposure and Auto White Balance algorithms specifically for Automotive Vision scenarios that is scalable to different embedded platforms co-existing with ISP framework

ADAS Applications in need for imaging algorithms

Surround View
Lane Departure Warning
Blind Spot Warning
Forward Collision Warning
Pedestrian Detection
Traffic Sign Recognition

RGB-IR for Automotive In-Cabin Application

Camera sensors supporting both RGB and IR feed are the new wave in the cabin monitoring application. This opens up the opportunity to utilize IR for machine vision and RGB for human vision application such as video conferencing. How ever most existing ISPs do not support RGB-IR. PathPartner supports leading ISPs with our proprietary pre processing algorithm for RGB-IR sensors.

PathPartner’s preprocessing algorithm for RGB-IR pixel array effectively does the extraction of RGB pixels and IR pixels separately and enables the ISP pipeline to enhance the image quality independently. The algorithm also takes care of removing the interference of IR on RGB data. These independently enhanced images (RGB and IR) enable the user to work on human vision and machine vision applications simultaneously.

Test Results

Supported Platforms

Platform's: TDAxx

Sensor: Omnivison

By submitting this form, you authorize PathPartner to contact you with further information about our relevant content, products and services. You may unsubscribe any time. We are committed to your privacy. For more details, refer our Privacy Policy

Back to Top