Help with a visual SLAM project with C++
Budget: €12 – €18 EUR
I need someone to help me with a stereo visual SLAM/ stereo visual Odometry project.
The subtasks in the project include among others:
1. Detection and description of ORB features.
2. Stereo matching/stereo calibration and rectification.
3. Generating a disparity map/depth map/point cloud.
4. Relative pose estimation and calculating the rotation matrix and translation vector using PnP method.
5. If the rotation matrix and translation vector are good enough, we can also start building a map and draw pose graph and optimize it.
6. The whole project should be implemented on ROS platform.
Your profile (MUST HAVE):
1. Must have a good understanding of the concept of visual SLAM.
2. You must have very good C++ programming skills, opencv, computer vision and ROS platform.
I have already done the project on the ROS platform but the resulting rotation matrix and translation vector make no sense so I need someone to help me troubleshot the error and correct any mistake I have done and I’ll provide you with the C++ code I have done so far so we can assume that if you have a good understanding of the subject, then it'll not take too much time to find and correct my mistakes.
The criteria to accept your work as a SUCCESS:
1. When you get/ help me get the rotation matrix and translation vector that make sense and match the ground truth with an acceptable error.
2. Clarify my misunderstanding regarding the concept of the code.
The subtasks in the project include among others:
1. Detection and description of ORB features.
2. Stereo matching/stereo calibration and rectification.
3. Generating a disparity map/depth map/point cloud.
4. Relative pose estimation and calculating the rotation matrix and translation vector using PnP method.
5. If the rotation matrix and translation vector are good enough, we can also start building a map and draw pose graph and optimize it.
6. The whole project should be implemented on ROS platform.
Your profile (MUST HAVE):
1. Must have a good understanding of the concept of visual SLAM.
2. You must have very good C++ programming skills, opencv, computer vision and ROS platform.
I have already done the project on the ROS platform but the resulting rotation matrix and translation vector make no sense so I need someone to help me troubleshot the error and correct any mistake I have done and I’ll provide you with the C++ code I have done so far so we can assume that if you have a good understanding of the subject, then it'll not take too much time to find and correct my mistakes.
The criteria to accept your work as a SUCCESS:
1. When you get/ help me get the rotation matrix and translation vector that make sense and match the ground truth with an acceptable error.
2. Clarify my misunderstanding regarding the concept of the code.