Computer vision specialist Arducam has launched a new camera module for the Raspberry Pi, but this time with a difference: it's a time-of-flight (ToF) sensor, designed to provide depth information at 30 frames per second.
Where Arducam's previous products have primarily focused on capturing visible light at ever-increasing resolutions, its latest Raspberry Pi-compatible camera module switches to active depth sensing through time-of-flight (ToF) measurement.
The camera combines a 940nm infrared vertical-cavity surface-emitting laser (VCSEL) with a 240×180-resolution image sensor, technically capable of running at 120 frames per second but limited to 30 frames per second on the Raspberry Pi. With a 70-degree field-of-vision (FoV), the sensor can capture depth data at up to 6.5 feet in "near mode" and 13 feet in "far mode."
Measuring just 38×38mm (around 1.5×1.5"), the camera connects to the Raspberry Pi's MIPI Camera Serial Interface (CSI) port plus the general-purpose input/output (GPIO) header for the laser's power and offers C, C++, and Python sample code for raw data capture and depth processing. Arducam's drivers also expose the sensor as a standard V4L2 camera, providing compatibility with computer vision and machine learning frameworks including OpenCV and TensorFlow.
The ToF sensor's release follows on the heels of the company's latest USB 3.0 camera sensor evaluation kit and ultra-high-resolution 108 megapixel camera sensor, which in turn serves as a higher-resolution alternative to the already-impressive Pi Hawk-eye camera at 64 megapixels — itself offered as a successor to an earlier lower-cost 16 megapixel autofocus camera module design.
The Arducam ToF Camera is now funding on Kickstarter, priced at $29.90 for backers as a discount from a planned retail price of $50.