11 skills found
jagennath-hari / RGBD 3DGS SLAMRGBD-3DGS-SLAM is a monocular SLAM system leveraging 3D Gaussian Splatting (3DGS) for accurate point cloud and visual odometry estimation. By integrating neural networks, it estimates depth and camera intrinsics from RGB images alone, with optional support for additional camera information and depth maps.
Mazhichaoruya / Object Detection And Location RealsenseD435Use the Intel D435 real-sensing camera to realize target detection based on the Yolov3 framework under the Opencv DNN framework, and realize the 3D positioning of the Objection according to the depth information. Real-time display of the coordinates in the camera coordinate system.ADD--Using Yolov5 By TensorRT model,AGX-Xavier,RealTime Object Detection
TakashiYoshinaga / IPhoneCinematicDepthTo3DThis is a sample C# project that extracts Depth and Color information from videos shot in iPhone's Cinematic mode and outputs each as separate videos, along with a sample Unity project for 3D playback of these videos.
yzn9961 / RGBD2PointCloud一个将Astra pro深度相机采集到的图像信息转换为三维点云的简单项目。项目中同时包括了Astra Pro RGBD相机标定、图像&视频采集的代码。A simple demon to convert the Depth&RGB information capture by Astra pro camera into 3D point cloud. (Photo&video capture and camera calibration code are also included)
grigala / 3DMMDepthFittingEfficient 3D Morphable Model Face Fitting using Depth Sensing Technologies
SuhailAhmadMir / An Intelligent EyeDeveloping a tool for the visually impaired people is not a recently emerged problem. But developing a computer aided tool is a still developing area. The aim of all these systems is to help the user in navigation without the help of a second person. There are several works using computer vision techniques. But there is no existing method that help to solve the all basic needs of blind person. This project presents a comprehensive scheme for reconstructing a three-dimensional (3D) model from a stereo camera via multi-view calibration. The depth information is useful so as to estimate the actual position of the target object. It is an essential parameter that can allow visualization from multiple perspectives. We provide an improved method for detecting objects and calculating the distance of these objects using stereo vision in real time. The stages involved include camera calibration, image rectifying, disparity calculation, and three dimensional reconstruction. Results show that objects located from 65cm up to 203cm are properly measured for their distances, with an average error of 3cm. The precision of the measurement also depends on the quality of the calibration.
Pirate-Emperor / DTM 3DDTM_3D for Moon Images is an advanced project developed by Pirate-Emperor that specializes in Depth Terrain Estimation for lunar landscapes. This project utilizes cutting-edge computer vision and deep learning techniques to estimate the depth information of terrain from 2D images of the Moon. DTM_3D for Moon Images is designed for such applications
connahKendrickMMU / Towards Real Time Facial Landmark Detection In Depth Data Using Auxiliary InformationModern facial motion capture systems employ a two-pronged approach for capturing and rendering facial motion. Visual data (2D) is used for tracking the facial features and predicting facial expression, whereas Depth (3D) data is used to build a series of expressions on a 3D face models. An issue with modern research approaches is the use of a single data stream that provides little indication of the 3D facial structure. We compare and analyse the performance of Convolutional Neural Networks (CNN) using visual, Depth and merged data to identify facial features in real-time using a Depth sensor. First, we review the facial landmarking algorithms and its datasets for Depth data. We address the limitation of the current datasets by introducing the Kinect One Expression Dataset (KOED). Then, we propose the use of CNNs for the single data stream and merged data streams for facial landmark detection. We contribute to existing work by performing a full evaluation on which streams are the most effective for the field of facial landmarking. Furthermore, we improve upon the existing work by extending neural networks to predict into 3D landmarks in real-time with additional observations on the impact of using 2D landmarks as auxiliary information. We evaluate the performance by using Mean Square Error (MSE) and Mean Average Error (MAE). We observe that the single data stream predicts accurate facial landmarks on Depth data when auxiliary information is used to train the network.
m-wahaha / Binocular 3D ReconstructionMainly used to compare SIFT and SURF in 3D reconstruction. Includes image preprocessing, feature extraction and matching, parallax and depth information, 3D reconstruction.
Yiren-20 / Stereo Vision With Binocular CamerasThis project focuses on the principles of a binocular stereo vision system, system calibration, image adjustment, image matching, depth information calculation, and point cloud generation, with the goal of completing 3D reconstruction based on binocular vision.
balamuruganky / Depthmap To PcdExtract Point ClouD (PCD) from Depth Map information from 3D camera (OpenNI and FreeNECT supported)
