Synchronized Dual Lens Stereo USB Camera 1.3MP HD 960P Webcam 3D VR Web Camera Module with 1/3 CMOS OV9715 Sensor Mini Industrial USB2.0 Web cam for Android,Linux,Windows,Raspberry PI

Does what it says. Gives a single video feed of the images laid side by side.

Camera came quick and integrated perfectly with ROS. Can be used with robotics and many other deployments.

It's a really neat camera. I had no problems getting it running on PC, hopefully Linux is similar. The PCB feels excellent and looks really well made. Image/video quality is better than you'd expect for sub full HD. So this has an ELP sticker and manual, looks like an ELP manufactured product, but the design dates back to 2021. A quick search shows that ELP has updated this recently with 4mp cameras and no longer sells the 1.3mp version. At the time of review, the prices between the two are very similar right here on Amazon, and because of that, I cannot say this is anything more than average. If prices shift, this might make more sense as it is a good product, just a bit old hat.

In terms of video streams or pseudo VR or stereo video to youtube this would work well but have not done this. My primary goal is to use this for CV based applications which it does seem to work well with opencv2 without initial tuning. With tuning it should be really good given the frames are synced. for CV: In the attached images shows a A4 paper with markings at 5/10/15/20/25cm and lines going from the middle of the page to the 25cm end points - about 45deg angle. For ML/vision the cameras need to be fairly aligned otherwise software compensation is required to re-tune the images. All of that is possible with tuning but the point being demonstrated is that the cameras are not fully identical to begin with. The differences in images when matching the center line (moving the camera left/right) shows that there is quite a bit of distortion or focal issues. Despite these misalignments CV was able to compensate and create a good disparity image. I look forward to fine-tuning this and 3d printing a housing to keep the circuits safe and to be used for object scanning Code for CV testing (most can be found online, but this one works with live video feed) Note: Formatting may be impacted, adjustment likely required #pip.exe install opencv-contrib-python import cv2 import numpy as np #pip.exe install matplotlib import matplotlib.pyplot as plt cam = cv2.VideoCapture(1, cv2.CAP_DSHOW) #cam.set(cv2.CAP_PROP_FRAME_WIDTH, 2560) #cam.set(cv2.CAP_PROP_FRAME_HEIGHT, 960) image = cam.read()[1] w = int(image.shape[1]/2) h = int(image.shape[0]) print('Width: ', w) print('Height: ', h) stereo = cv2.StereoBM.create(numDisparities=16*3, blockSize=5) #stereo.setMinDisparity(0) #stereo.setTextureThreshold(10) stereo.setSpeckleWindowSize(300) #stereo.setPreFilterSize(9) stereo.setSpeckleRange(50) stereo.setUniquenessRatio(10) while(True): image = cv2.rotate(cam.read()[1], cv2.ROTATE_180) iml_color = image[0:h, 0:w] imr_color = image[0:h, w:w+w] iml_bw = cv2.cvtColor(iml_color, cv2.COLOR_RGB2GRAY) imr_bw = cv2.cvtColor(imr_color, cv2.COLOR_RGB2GRAY) #apply blur #iml_bw = cv2.blur(iml_bw,(1,1)) #imr_bw = cv2.blur(imr_bw,(1,1)) cv2.imshow("left", iml_color) cv2.imshow("right", imr_bw) disparity = stereo.compute(iml_bw, imr_bw) disparity_scale = disparity cv2.imshow("dsp", disparity_scale.astype('uint8')) print(f"Range: {np.min(disparity)} <-> {np.max(disparity)}") disparityImg = cv2.normalize(src=disparity, dst=None, alpha=00, beta=255, norm_type=cv2.NORM_MINMAX, dtype=cv2.CV_8UC1) disparityImg = cv2.applyColorMap(disparityImg, cv2.COLORMAP_JET) cv2.imshow("depth", disparityImg) if cv2.waitKey(1) & 0xFF == ord('q'): break cam.release() cv2.destroyAllWindows()

This is a simple, bare-bones SBS (side-by-side) 3D USB video camera that takes excellent high-quality 3D photos or video using any video capture app. You can also use it as a 3D web cam if you like. It doesn't come with any bells and whistles, or a case for that matter, just a simple circuit board with a camera on each end, as seen in the product photos. Just plug it into your computer using the supplied USB cable. If you use a Mac computer, just open Quicktime and start a New Movie Recording. You can choose the 3D USB camera from the dropdown next to the record button. If you use a PC, you can use any of the any of the free video capture apps that are available. AMCap is a decent one that is suggested in the user manual.