Search Results

In this work, we outline a process for traffic light detection in the context of autonomous vehicles and driver assistance technology features. For our approach, we leverage the automatic annotations from virtually generated data of road scenes. Using the automatically generated bounding boxes around the illuminated traffic lights themselves, we trained an 8-layer deep neural network, without pre-training, for classification of traffic light signals (green, amber, red). After training on virtual data, we tested the network on real world data collected from a forward facing camera on a vehicle. Our new region proposal technique uses color space conversion and contour extraction to identify candidate regions to feed to the deep neural network classifier. Depending on time of day, we convert our RGB images in order to more accurately extract the appropriate regions of interest and filter them based on color, shape and size. These candidate regions are fed to a deep neural network.

This work describes a single camera based object distance estimation system. As technology on vehicles is constantly advancing on the road to autonomy, it is critical to know the locations of objects in 3D space for safe behavior of the vehicle. Though significant progress has been made on object detection in 2D sensor space from a single camera, this work additionally estimates the distance to said object without requiring stereo vision or absolute knowledge of vehicle motion. Specifically, our proposed system is comprised of three modules: vision based ego-motion estimation, object-detection, and distance estimation. In particular, we compensate for the vehicle ego-motion by using pin-hole camera model to increase the accuracy of the object distance estimation.

Heart rate is one of the most important biological features for health information. Most of the state-of-the-art heart rate monitoring systems rely on contact technologies that require physical contact with the user. In this paper, we discuss a proof-of-concept of a non-contact technology based on a single camera to measure the user’s heart rate in real time. The algorithm estimates the heart rate based on facial color changes. The input is a series of video frames with the automatically detected face of the user. A Gaussian pyramid spatial filter is applied to the inputs to obtain a down-sampled high signal-to-noise ratio images. A temporal Fourier transform is applied to the video to get the signal spectrum. Next, a temporal band-pass filter is applied to the transformed signal in the frequency domain to extract the frequency band of heart beats. We then used the dominant frequency in the Fourier domain to find the heart rate.