"Visual SLAM fourteen say: From Theory to Practice" describes the visual system SLAM (simultaneous localization and mapping) required basic knowledge and core algorithms, both theoretical foundations of mathematics, such as rigid body motion in three-dimensional space, nonlinear optimization, further comprising a computer vision algorithms, such as multi-view geometry, loop detection. In addition, it also provides a lot of example code readers for study and research to more in-depth grasp of the content.
"Computer Vision in multi-view geometry" comprehensive introduction nearly 10 years to develop the calculation method based on geometric and mathematical foundations of computer vision. In addition to the above, in which the multi-camera view geometry and calculation methods, the reader is worth attention.
"Visual SLAM fourteen From theory to practice" high-definition PDF, 399 pages, table of contents with bookmarks, color with pictures, text can be copied. Complete source code.
"Computer Vision in multi-view geometry," the Chinese version, 484;
"Computer Vision multi-view geometry (2nd Edition)" in English, 673;
Complete source code + illustrations.
Link: https: //pan.baidu.com/s/17VeMzDw5DY91tAyg15vaQw
extraction code: pglx
copy the contents of this open Baidu network disk phone App, the operation more convenient oh
"Visual SLAM fourteen say: from theory to practice" self-study materials can be used as the entry of researchers interested in SLAM can also be used as college undergraduate or graduate curriculum materials related to the use of SLAM.
The current price-performance computer greatly improved, so that people conditional use more cameras in the visual system to take advantage of redundant information exchange system robustness to noise. The noise of the robustness of the system has been a bottleneck practical computer vision system, a possible solution to this problem is to: improve the resolution of the camera, multi-camera methods and statistics in recent years introduced a large number of robust optimization algorithm (book many chapters also described). "Computer Vision in multi-view geometry" on China specializing in computer vision research readers have a better reference value. In addition, persons engaged in research related to the field of mathematics is also worth reading.