title
stringlengths 1
544
โ | parent
stringlengths 0
57
โ | created
stringlengths 11
12
โ | editor
stringclasses 1
value | creator
stringclasses 4
values | edited
stringlengths 11
12
โ | refs
stringlengths 0
536
โ | text
stringlengths 1
26k
| id
stringlengths 32
32
|
|---|---|---|---|---|---|---|---|---|
Realtime Rendering
|
Rendering Usages
|
Dec 6, 2021
|
Alan Jo
|
Alan Jo
|
Oct 5, 2022
|
14423ba9d05f485eb610848b8175086f
|
||
Rendering Engine
|
Rendering Usages
|
Nov 26, 2021
|
Alan Jo
|
Alan Jo
|
Dec 6, 2021
|
### Rendering Engines
|Title|
|:-:|
|[Filament](https://texonom.com/filament-2ceb8840426e4e7eb3143c162ee1b7e9)|
|
b697c66272d344a08354189623636d51
|
|
Filament
|
Rendering Engines
|
Nov 26, 2021
|
Alan Jo
|
Alan Jo
|
Nov 26, 2021
|
[undefined](https://github.com/google/filament)
|
### Cross platform rendering engine from google
Android, iOS, Windows, Linux, macOS, and WebGL2
|
2ceb8840426e4e7eb3143c162ee1b7e9
|
Metaverse
|
Simulations
|
Nov 23, 2020
|
Alan Jo
|
Alan Jo
|
Apr 4, 2022
|
[Game](https://texonom.com/game-66ff5030854947d0b23ae4b862f40590) [BCI](https://texonom.com/bci-e7b6d2a55d19435b9d3f8990441cab9e) [Meta Platforms](https://texonom.com/meta-platforms-d7c44ce8da1e4eec91d1b11b9e5c2f90)
|
### ๋ฉํ๋ฒ์ค๊ฐ ๊ฒ์์์ ๋ถ๋ฆฌ๋ ์ด์ ๋ ๋จ์ํ ์์ ๋ ๋์ ๊ฒ์์์ ์ถ๋ฐํด, ์ฐฝ์์ ์ธ ์ด์ฉ์๋ค์ด ์๋ก์ด ์ปจํ
์ธ ๋ฅผ ๋ง๋ค ์ ์๋ ํ๋ซํผ ํํ๋ฅผ ๊ฐ์ถ์๊ธฐ ๋๋ฌธ
๋ฉํ๋ฒ์ค์์ ๊ฒ์์ ์์. ๋ฉํ๋ฒ์ค์ ๊ฐ์ฅ ์ค์ํ ์์ญ์ ์ฌ๊ต
> [Metaversatility :: Home](http://metaversatility.com/)
|
ce545edddf6d458d94e98c95a92b9b83
|
Virtual Twin
|
Simulations
|
May 26, 2021
|
Alan Jo
|
Alan Jo
|
Jun 7, 2021
|
์ปดํจํฐ์ ํ์ค ์ ์ฌ๋ฌผ์ ์๋ฅ์ด๋ฅผ ๋ง๋ค๊ณ , ํ์ค์์ ๋ฐ์ํ ์ ์๋ ์ํฉ์ ์ปดํจํฐ๋ก ์๋ฎฌ๋ ์ด์
ํจ์ผ๋ก์จ ๊ฒฐ๊ณผ๋ฅผ ๋ฏธ๋ฆฌ ์์ธกํ๋ ๊ธฐ์
|
5c46be2b60c2407790c1b8b16dcda533
|
|
Computer Vision Library
|
Computer Vision Usages
|
May 13, 2021
|
Alan Jo
|
Alan Jo
|
May 16, 2023
|
### Computer Vision Libraries
|Title|
|:-:|
|[OpenCV](https://texonom.com/opencv-c9ddfabf30094b25a742106592f55d6d)|
|
e68fc326f8ba489cbd6b90628a1887f9
|
|
Computer Vision Sensor
|
Computer Vision Usages
|
May 13, 2021
|
Alan Jo
|
Alan Jo
|
Jun 28, 2021
|
### Computer Vision Sensors
|Title|
|:-:|
|[LIDAR](https://texonom.com/lidar-944e820b845645fc8ca1ba6f61a8119e)|
### Computer Vision Sensor Notion
|Title|
|:-:|
|[Odometry](https://texonom.com/odometry-d54eb30e7e29495e84e326438662a0e7)|
|
a925bbb41c6b4e8c9ac479d03caa8014
|
|
Computer Vision Target
|
Computer Vision Usages
|
Jun 28, 2021
|
Alan Jo
|
Alan Jo
|
Jun 6, 2023
|
### Computer Vision Targets
|Title|
|:-:|
|[Object Detection](https://texonom.com/object-detection-5f12617d24ee490ea11584ba74b110ed)|
|[Depth Estimation](https://texonom.com/depth-estimation-2f7e039fb9584b50b76bbb0a98bce451)|
|[Corner Detection](https://texonom.com/corner-detection-36dc7afee7414060bd66c1003898836e)|
|[Shape Modeling](https://texonom.com/shape-modeling-84a770ac7869447290dfcf37ea83e465)|
|[Behavior Recognition](https://texonom.com/behavior-recognition-61a36691a014462dad40d6a3e78a4ad2)|
|[Data Annotation](https://texonom.com/data-annotation-867099227cdf49c0869e027106b545d8)|
|[Image Segmentation](https://texonom.com/image-segmentation-eeefd926f87f457eba4034698997d86a)|
|[Edge Detector](https://texonom.com/edge-detector-7c0f6ccc9f6f49a380dbcdf347883eed)|
|[Motion Perception](https://texonom.com/motion-perception-73bdb8160bf6470d83002d265f8306b9)|
|[Document Scanning](https://texonom.com/document-scanning-fed2929116884dcea3e3023105fbc6b5)|
|[Image Captioning](https://texonom.com/image-captioning-b1fa39da247648b5a514f9f4b994e4ab)|
|[View Synthesis](https://texonom.com/view-synthesis-fb0b95a897f34fdba500a8b21aee82d0)|
|
74fdd39f373e4502987583cbe21e8036
|
|
Non Line of Sight
|
Computer Vision Usages
|
Apr 6, 2023
|
Alan Jo
|
Alan Jo
|
Apr 6, 2023
|
ํผ์ฝ๋ฏธํฐ ํ๋ ์ ์นด๋ฉ๋ผ ํ์ํด์ ๋ผ์ด๋ค๋ณด๋ค ๋น์๋ฏ
๊ทธ๋ฆฌ๊ณ ์๋์ฐจ๊ฐ์๊ฑด ์ง๋ ์ธ์ ์ ํ๋ ๋์ค๊ธฐ๊ฐ ํ๋ค๋ฏ ์ ์ง์ํฉ์์๋ ์๋์ค๋๋ฐ
> [Non-Line-of-Sight Imaging | Edmund Optics](https://www.edmundoptics.com/knowledge-center/trending-in-optics/non-line-of-sight-imaging/)
> [Non-line-of-sight Imaging with Partial Occluders and Surface Normals | TOG 2019](https://www.youtube.com/watch?v=nVoEZuFFMzA)
> [David B. Lindell | Confocal non-line-of-sight imaging based on the light-cone transform](https://davidlindell.com/publications/nlos-lct)
|
22932d095eeb4bd987394eda7e909696
|
|
View Synthesis
|
Computer Vision Usages
|
Aug 17, 2022
|
Alan Jo
|
Alan Jo
|
May 16, 2023
|
> [https://www.youtube.com/watch?v=7iy0WJwNmv4&feature=youtu.be](https://www.youtube.com/watch?v=7iy0WJwNmv4&feature=youtu.be)
|
3106a9864bdb46a7be7ac5fd892ea13b
|
|
OpenCV
|
Computer Vision Libraries
|
Nov 19, 2019
|
Alan Jo
|
Alan Jo
|
May 16, 2023
|
### OpenCV Notion
|Title|
|:-:|
|[Open CV T-API](https://texonom.com/open-cv-t-api-4b33a833917e40288d19a484209f117d)|
|[Open CV IPP](https://texonom.com/open-cv-ipp-1b1e07ef7d4247f1831ccd5e71f74ed4)|
|[Open CV hierarcy](https://texonom.com/open-cv-hierarcy-a5a8ad117fb04006983b3d2bb460be72)|
|[Open CV History](https://texonom.com/open-cv-history-3ecc2ad4c0fb4a42b9746f4879499d1f)|
### OpenCV Usage
|Title|
|:-:|
|[OpenCV Type](https://texonom.com/opencv-type-7263db318af841209630036aa44dd257)|
|[Open CV Window](https://texonom.com/open-cv-window-4720fb20595944a39d817b4e1ddc2734)|
|[Open CV Tick](https://texonom.com/open-cv-tick-87b6863dde694862a30c2f7c1c8c8a84)|
|[Open CV Filtering](https://texonom.com/open-cv-filtering-d5b2a9d648ad489db395480a900c7cc3)|
|[Open CV Event](https://texonom.com/open-cv-event-71c15f9be2ca45399aab090679116221)|
|[Open CV Draw](https://texonom.com/open-cv-draw-d2a8fe7ec315439fa6bcdfb8964e4d48)|
|[Open CV Implementation](https://texonom.com/open-cv-implementation-a8866ff33d2047908acfb6bbd92a179e)|
> [[5.1.4] ํธ๋๋ฐ๋ฅผ ์ด์ฉํ ์์์ ๋ฐ๊ธฐ ์กฐ์ ](https://www.youtube.com/watch?v=znXuTOLs-4c&list=PLr4Kym6mXRgJhObgeJQyBk2ZJ3FXGJK_a&index=6)
> [bsdnoobz/opencv-code](https://github.com/bsdnoobz/opencv-code/blob/master/eye-tracking.cpp)
> [adamian98/pulse](https://github.com/adamian98/pulse)
|
c9ddfabf30094b25a742106592f55d6d
|
|
Open CV hierarcy
|
OpenCV Notion
|
Nov 19, 2019
|
Alan Jo
|
Alan Jo
|
Apr 16, 2023
|
each module has dll file and lib file (dynamic and static)
# core module
- main class and functions
# highgui module
- video export function
# Others
[dnn module](https://texonom.com/dnn-module-9156adc3304845d5a883a6a3e7cadd02)
[ml module](https://texonom.com/ml-module-185c6f2d0e77434c87f5b08219625bad)
|
a5a8ad117fb04006983b3d2bb460be72
|
|
Open CV History
|
OpenCV Notion
|
Nov 19, 2019
|
Alan Jo
|
Alan Jo
|
Apr 16, 2023
|
From Intel IPL to open source
|
3ecc2ad4c0fb4a42b9746f4879499d1f
|
|
Open CV IPP
|
OpenCV Notion
|
Nov 19, 2019
|
Alan Jo
|
Alan Jo
|
Apr 21, 2021
|
- GPU Usage more
|
1b1e07ef7d4247f1831ccd5e71f74ed4
|
|
Open CV T-API
|
OpenCV Notion
|
Apr 21, 2021
|
Alan Jo
|
Alan Jo
|
Apr 21, 2021
|
4b33a833917e40288d19a484209f117d
|
||
dnn module
|
Open CV hierarcy
| null | null | null | null | null |
9156adc3304845d5a883a6a3e7cadd02
|
|
ml module
|
Open CV hierarcy
| null | null | null | null | null |
185c6f2d0e77434c87f5b08219625bad
|
|
Open CV Draw
|
OpenCV Usage
|
Nov 19, 2019
|
Alan Jo
|
Alan Jo
|
Apr 16, 2023
|
- thinkness extend size
# line
- line (inputoutput, point, point, scalar(color), thickness, linetype, int shift bit > 0)
> - shift move position only
- arrowedLine(same, double tip length = 0.1 size of tip)
- drawMarker(img, pt, color, markertype, size, thick, linetype)
# line Figure
- rectengular(line arguments) or (two point to rect)
> - if thinkness is minus โ fill rect
- circle(img, center, radius, color, thickness, linetype, shift)
- elipse(img, center, axes, angle, startAngle, endAngle, color, thickness, lineType, shift)
- polylines(img, pts(vector<points >usually), bool closed, color, thickness, linetype, shift)
# Text
- putText(img, text, org(left bottom), fibtface from hersheyFonts, fontscale, color, thickness, linetype, bottomeleftorgin if false โ leftTop)
- getTextSize(text, fontFace, fontScale, thickness, baseline) - do not include all size - little small
|
d2a8fe7ec315439fa6bcdfb8964e4d48
|
|
Open CV Event
|
OpenCV Usage
|
Nov 19, 2019
|
Alan Jo
|
Alan Jo
|
Apr 16, 2023
|
# Keyboard
- keywait
- WaitKeyEx - other insert or function key
# Mouse
- setMouseCallback -(winname, onmouse, userdata pointer)
>
# Trackbar
- createTrackbar(barname, winname, default value pointer, count, callback, )userdata
- gettrackbar position
- settracbarposition
|
71c15f9be2ca45399aab090679116221
|
|
Open CV Filtering
|
OpenCV Usage
|
Nov 21, 2019
|
Alan Jo
|
Alan Jo
|
Apr 16, 2023
|
### mast = window = kernel (usually 3x3)
- usually black and white cv_u8
- execute for not zery
# Kind
- (gausian) blur <> sharpening - used for out focusing
- embosing โ edge
- remove noise = remove gausian noise from camera
> - **edge-preserving noise removal filter - bilateral filter**
> - median filering - remove salt & pepper noise
# function
- SetTo(inputarrary value, inputarrary mask = noaarrary)
- copyTo(output img, input mask)
-
# Anchor Point
# Correlation and Convolution
- correlation
> - mast
- convolution
> - multiply after Origin symmetry
usually same in cv because in cv filter mask is symmetry
|
d5b2a9d648ad489db395480a900c7cc3
|
|
Open CV Implementation
|
OpenCV Usage
|
Jul 29, 2021
|
Alan Jo
|
Alan Jo
|
Apr 12, 2023
|
### Open CV Implementations
|Title|
|:-:|
|[C++ OpenCV](https://texonom.com/c-opencv-adcd09674dc04c3aaf057b82c6191000)|
|[Python OpenCV](https://texonom.com/python-opencv-76b55880579547bab1c5ccc0e723a046)|
|[Javascript OpenCV](https://texonom.com/javascript-opencv-87365f4375d64009a6414b609c89755d)|
|
a8866ff33d2047908acfb6bbd92a179e
|
|
Open CV Tick
|
OpenCV Usage
|
Nov 23, 2019
|
Alan Jo
|
Alan Jo
|
Apr 16, 2023
|
```typeint64 getTickCount(void)return - now Tick Count```
```typedouble getTickFrequency(void)return - System Tick Freuency```
```typeClass TickMeter start(); stop(); double getTimeMicro ... int64 getCounter```
|
87b6863dde694862a30c2f7c1c8c8a84
|
|
Open CV Window
|
OpenCV Usage
|
Nov 19, 2019
|
Alan Jo
|
Alan Jo
|
Apr 16, 2023
|
- if made by WINDOW_AUTOSIZE โ cannot change window size
- imshow() and waitKey is almost come together
|
4720fb20595944a39d817b4e1ddc2734
|
|
OpenCV Type
|
OpenCV Usage
|
Apr 3, 2022
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
### OpenCV Types
|Title|
|:-:|
|[OpenCV Class](https://texonom.com/opencv-class-fcfe8a2b74434cb59f4de492ec73f8f3)|
|[OpenCV Interface Class](https://texonom.com/opencv-interface-class-973c198295194ef09de7cab123167bc7)|
|[OpenCV Template Class](https://texonom.com/opencv-template-class-416a8c70e7e6402c88322b3c8ef22cf6)|
|
7263db318af841209630036aa44dd257
|
|
C++ OpenCV
|
Open CV Implementations
|
Jul 29, 2021
|
Alan Jo
|
Alan Jo
|
Jul 29, 2021
|
adcd09674dc04c3aaf057b82c6191000
|
||
Javascript OpenCV
|
Open CV Implementations
|
Jul 29, 2021
|
Alan Jo
|
Alan Jo
|
Jul 29, 2021
|
> [OpenCV: Using OpenCV.js](https://docs.opencv.org/3.4/d0/d84/tutorial_js_usage.html)
|
87365f4375d64009a6414b609c89755d
|
|
Python OpenCV
|
Open CV Implementations
|
Jul 29, 2021
|
Alan Jo
|
Alan Jo
|
May 28, 2023
|
```typepip install opencv-python```
### Python OpenCV Usages
|Title|
|:-:|
|[Python OpenCV Stitching](https://texonom.com/python-opencv-stitching-1ec0f97afb5b43b890722662e9eee785)|
### Not conda
[opencv-python intellisense is not working under conda](https://github.com/microsoft/vscode-python/issues/99)
|
76b55880579547bab1c5ccc0e723a046
|
|
Python OpenCV Stitching
|
Python OpenCV Usages
|
May 28, 2023
|
Alan Jo
|
Alan Jo
|
May 28, 2023
|
> [๐งฉImage Stitching ํ๋ก์ ํธ](https://velog.io/@davkim1030/Image-Stitching)
|
1ec0f97afb5b43b890722662e9eee785
|
|
OpenCV Class
|
OpenCV Types
|
Apr 3, 2022
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
### OpenCV Classes
|Title|
|:-:|
|[Open CV Mat](https://texonom.com/open-cv-mat-ca5909c92a5d49ae82e34123e0b57951)|
|[OpenCV VideoWriter](https://texonom.com/opencv-videowriter-dddfa63758a444af9c159e61d09be81b)|
|[OpenCV FileStorage](https://texonom.com/opencv-filestorage-a1fcdad49e324bee8e32aabbc038a258)|
|[OpenCV VideoCapture](https://texonom.com/opencv-videocapture-879b1c8ced6c4f6e97e4ad75cff86e2c)|
|[OpenCV Vec](https://texonom.com/opencv-vec-43a311696d0f43a5930d9cd0d2ba56f5)|
|[OpenCV RotatedRect](https://texonom.com/opencv-rotatedrect-11026fb43e2649e08c78c90868537e2d)|
|[OpenCV Range](https://texonom.com/opencv-range-7ba3d683be1248efac3dc1a0fa8a358d)|
|
fcfe8a2b74434cb59f4de492ec73f8f3
|
|
OpenCV Interface Class
|
OpenCV Types
|
Apr 3, 2022
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
### OpenCV Interface Classes
|Title|
|:-:|
|[OpenCV inputArray ](https://texonom.com/opencv-inputarray-e2702656954f43b7afb7dc5598856bd5)|
|[OpenCV OutputArray](https://texonom.com/opencv-outputarray-e94872801feb47c99c125d85c093e597)|
|[inputOutputArray](https://texonom.com/inputoutputarray-060d047cb62d4b21969315a4ae9ac7e5)|
|
973c198295194ef09de7cab123167bc7
|
|
OpenCV Template Class
|
OpenCV Types
|
Apr 3, 2022
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
### OpenCV Template Classes
|Title|
|:-:|
|[OpenCV Point_](https://texonom.com/opencv-point-5ddd7f31bf0e404e855ac9f02d3f5893)|
|[OpenCV Size_](https://texonom.com/opencv-size-1992ca2dc472424d96d152c3272acb3b)|
|[OpenCV Rect_](https://texonom.com/opencv-rect-57eee0c38a0f48fb87aa11d2b93dc11f)|
|
416a8c70e7e6402c88322b3c8ef22cf6
|
|
Open CV Mat
|
OpenCV Classes
|
Nov 19, 2019
|
Alan Jo
|
Alan Jo
|
Mar 1, 2023
|
- Mat::empty needed
- Save
> - vector field
> - point cloud
> - histogram
> - ... usually 2d video data
- mat(vector) return 1 col โ need to .t
- MatExpr
> [What is the purpose of cv::MatExpr?](https://stackoverflow.com/questions/39964583/what-is-the-purpose-of-cvmatexpr)
# Create
```typeMat((row, col), type, s(B, G, R) or (data, (padding byte)))```
- eye
- one
- zero
- create, setTo (init)
[OpenCV Mat Edit](https://texonom.com/opencv-mat-edit-af680cedd99b4b3fb325b4bc4fcefff0)
# Copy
= is shallow copy (share pixel data)
> - ~mat inverts pixcel data
- copyTo, clone is deep copy
# Get
- roi(region of interest) = Rect or (rowRange, colRange)
- colRange - shallow
- rowRange- shallow
- row - shallow
- col - shallow
### Pixel
- at<uchar>(**y, x**) - **Slow**
- ptr<>(y) return start pointer uchar* can access by p[i] - **Faster**
- MatIterator - **Slow**
> - mat::begin() - first pixel location
> - mat::end() - next of final pixel location
# Normalization
- convertTo - data by alpha and beta
- reshape - new row or or channel
- resize - new row and initialize
- pop_back - nelem remove row count
# Data
- Scalar sum
- Scalar mean
- minMAxLoc(ation)
|
ca5909c92a5d49ae82e34123e0b57951
|
|
OpenCV FileStorage
|
OpenCV Classes
|
Nov 21, 2019
|
Alan Jo
|
Alan Jo
|
Mar 1, 2023
|
- operator [] return FileNode Object
- << to save
- >> to read
- should be declare and assign simultaniously
# open
- open(name, flag(read or write , appedn, memory(read or write)), encoding)
- isopend
- release - release data in memory buffer
# Read
- fs[] = file node object
|
a1fcdad49e324bee8e32aabbc038a258
|
|
OpenCV Range
|
OpenCV Classes
|
Nov 20, 2019
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
7ba3d683be1248efac3dc1a0fa8a358d
|
||
OpenCV RotatedRect
|
OpenCV Classes
|
Nov 20, 2019
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
- has boundingRect
|
11026fb43e2649e08c78c90868537e2d
|
|
OpenCV Vec
|
OpenCV Classes
|
Nov 20, 2019
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
- Matx ์์
- Matx 1์ฐจ์ ๊ตฌํ โ iterator
- typedef vec3b (Vec<uchar, 3>), ...
>
> [OpenCV Scalar](https://texonom.com/opencv-scalar-87f3281fed5e4168835afaaa23c5e957)
>
# val
|
43a311696d0f43a5930d9cd0d2ba56f5
|
|
OpenCV VideoCapture
|
OpenCV Classes
|
Nov 20, 2019
|
Alan Jo
|
Alan Jo
|
Mar 1, 2023
|
- member variable is all protect โ use method
- constructor โ open()
# open
- constructor(name, video api)
> - constructor(0) means construct and open default camera
- bool open (index, camera api)
> - index = camera_id + domain_offset_id
- bool isOpend() const
# Realese
- realese() - called by destructor
- read() == >> do same - from first frame (grab() + retrieve())
- int var = cvRound(double get(videocapture property id))
- bool set(prop id, double)
|
879b1c8ced6c4f6e97e4ad75cff86e2c
|
|
OpenCV VideoWriter
|
OpenCV Classes
|
Nov 20, 2019
|
Alan Jo
|
Alan Jo
|
Mar 1, 2023
|
- forcc(Four Codec Code) makes forcc code member method
- constructor with no argument
# Write
- open(file, fxx, fps, framesize, iscolor)
- write(frame per frame) << - size should be same
- release() - called by destructor
|
dddfa63758a444af9c159e61d09be81b
|
|
OpenCV Mat Edit
|
Open CV Mat
| null | null | null | null | null |
- setTo (mat, mask)
> - noArray() = Mat()
> [์ด๋ฏธ์ง์ ๊ธฐํํ์ ๋ณํ - gramman 0.1 documentation](https://opencv-python.readthedocs.io/en/latest/doc/10.imageTransformation/imageTransformation.html)
# Color
- normalize
> - normalize norm
# Brightness
- template inline uchar saturate_cast<uchar> (int v)
# Histogram
### Graph of same Pixel data count
- streching
> just stretch 1 dimension
$$H(g) = \Sigma_{0<i<g}{h(i)}$$
- equalization
> take interval as how much thy are
|
af680cedd99b4b3fb325b4bc4fcefff0
|
OpenCV Scalar
|
OpenCV Vec
| null | null | null | null | null |
- Vec inheritence
- 4 double val
|
87f3281fed5e4168835afaaa23c5e957
|
inputOutputArray
|
OpenCV Interface Classes
|
Nov 20, 2019
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
060d047cb62d4b21969315a4ae9ac7e5
|
||
OpenCV inputArray
|
OpenCV Interface Classes
|
Nov 19, 2019
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
- interface class
InputArray ํ์
์ ๋ฐ๋๋ก ์ค๋ช
๋์ด ์๋ค๋ฉด Mat ํด๋์ค ๊ฐ์ฒด ๋๋ vector<T> ํ์
์ ๋ณ์๋ฅผ ์ ๋ฌํ๋ ํํ๋ก ์ฝ๋๋ฅผ ์์ฑํด์ผ ํฉ๋๋ค. ๋ง์ฝ InputArray ํ์
์ ํจ์ ์ธ์์ ๋น์ด ์๋ ํ๋ ฌ์ ์ ๋ฌํ๋ ค๋ฉด ํจ์ ์ธ์์ noArray() ๋๋ Mat()
make function variable input array
> inside function you have to change it go .getMat
|
e2702656954f43b7afb7dc5598856bd5
|
|
OpenCV OutputArray
|
OpenCV Interface Classes
|
Nov 20, 2019
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
์ฌ์ฉ ํธ์์ฑ์ ์ํ๊ตฌํ์ฒด
- input array inheritance
- create() fucntion and output
|
e94872801feb47c99c125d85c093e597
|
|
OpenCV Point_
|
OpenCV Template Classes
|
Nov 20, 2019
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
5ddd7f31bf0e404e855ac9f02d3f5893
|
||
OpenCV Rect_
|
OpenCV Template Classes
|
Nov 20, 2019
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
57eee0c38a0f48fb87aa11d2b93dc11f
|
||
OpenCV Size_
|
OpenCV Template Classes
|
Nov 20, 2019
|
Alan Jo
|
Alan Jo
|
Apr 3, 2022
|
1992ca2dc472424d96d152c3272acb3b
|
||
Odometry
|
Computer Vision Sensor Notion
|
Jun 28, 2021
| null | null | null |
์๊ฐ์ ๋ฐ๋ฅธ ์์น ๋ณํ๋ฅผ ์ถ์ ํ๊ธฐ ์ํด ๋ชจ์
์ผ์์ ๋ฐ์ดํฐ๋ฅผ ์ฌ์ฉํ๋ ๊ฒ
|
d54eb30e7e29495e84e326438662a0e7
|
|
LIDAR
|
Computer Vision Sensors
|
Jun 7, 2020
| null | null | null |
944e820b845645fc8ca1ba6f61a8119e
|
||
Behavior Recognition
|
Computer Vision Targets
|
Jun 28, 2021
|
Alan Jo
|
Alan Jo
|
Jun 28, 2021
|
61a36691a014462dad40d6a3e78a4ad2
|
||
Corner Detection
|
Computer Vision Targets
|
Mar 23, 2023
|
Alan Jo
|
Alan Jo
|
Mar 28, 2023
|
### Only
- If edge, no change along the edge direction
- if corner, significant change in all directions
### Corner Detection Notion
|Title|
|:-:|
|[Harris corner detector](https://texonom.com/harris-corner-detector-4c7d6a569a424acba2dcd39880fa3410)|
|[Multi-scale Blob Detection](https://texonom.com/multi-scale-blob-detection-fbe1995acb4042d9a96608fba167afe9)|
|
36dc7afee7414060bd66c1003898836e
|
|
Data Annotation
|
Computer Vision Targets
|
Sep 22, 2020
|
Alan Jo
|
Alan Jo
|
Jun 28, 2021
|
์ ํํ ์ธ์๋ฅ ์ด ์ค์
๊ฐ์ฒด๋ณด๋ค ํฌ๊ฒ ๊ทธ๋ฆฌ๋ ๊ฑฐ ๋ฌธ์ ๊ฐ ์ ๊ณ
๊ฐ์ฒด๋ณด๋ค ์๊ฒ ๊ทธ๋ฆฌ๋ฉด ๋ฌธ์ ๊ฐ ๋๋ค
๊ฐ์ฒด๊ฐ ๋ฌด์์ธ๊ฐ์ ๋ํ์ฌ ํ์งํ ๊ฐ์ ๊ฒฝ์ฐ์ ์์ฃผ ํฐ ๋ฌธ์ ๊ฐ ๋ ์ ์๋ค
๊ทธ ๋ฐ์ดํฐ๊ฐ ๊ธฐ์ค ๋ฐ์ดํฐ๋ก์ hdmap ์ ์ฌ์ฉ๋๋ ค๋ฉด ๋ฐ๋ณต์ ๋ฐ๋ณต์ ๊ฐํ ๊ฒ์๊ฐ ํ์
Data Quality Control
- ๋ฐ์ด๋ฉ ๋ฐ์ค padding ํฝ์
์ ์ด๋๊น์ง ํ์ฉํ๋ค
|
867099227cdf49c0869e027106b545d8
|
|
Depth Estimation
|
Computer Vision Targets
|
Nov 10, 2021
|
Alan Jo
|
Alan Jo
|
Apr 27, 2023
|
### One Image is inherently ambiguous
We need multi view geometry
### Depth Estimations
|Title|
|:-:|
|[NeRF](https://texonom.com/nerf-199b304a6e3a4f23b1f749c0b380d03e)|
|[Monocular Depth Estimation](https://texonom.com/monocular-depth-estimation-d776b0cc5eeb4288841644a89e2b7b23)|
|[LLFF](https://texonom.com/llff-0fb92745f0174cd59a9cb7085b335e01)|
|[DinoV2](https://texonom.com/dinov2-12e07337f08f47ad825a1c7a51e8157b)|
|[Neuralangelo](https://texonom.com/neuralangelo-0ac52e6e862f4a2d99c53d5606f57cde)|
### Depth Estimation Notion
|Title|
|:-:|
|[Error Map](https://texonom.com/error-map-5b277183ed644029adcbe0297fdc9c7d)|
|[Pinhole Camera Model](https://texonom.com/pinhole-camera-model-681f008c19d2456cb107e9a225a10d3a)|
|[Triangulation](https://texonom.com/triangulation-086333beed5f497786fbbf8658eb4aa1)|
|[Two-view Geometry](https://texonom.com/two-view-geometry-fe5db6a8c36e4f99abae2cb123ed1d6b)|
|
2f7e039fb9584b50b76bbb0a98bce451
|
|
Document Scanning
|
Computer Vision Targets
|
May 16, 2023
|
Alan Jo
|
Alan Jo
|
May 16, 2023
|
### Document Scanning Tools
|Title|
|:-:|
|[JSCanify](https://texonom.com/jscanify-9f2fccb78db04269931b4b11b3aebfc6)|
|
fed2929116884dcea3e3023105fbc6b5
|
|
Edge Detector
|
Computer Vision Targets
|
Apr 23, 2023
|
Alan Jo
|
Alan Jo
|
Apr 23, 2023
|
[Derivative Filter](https://texonom.com/derivative-filter-1c54bd4735874bfabc93b6946b019936)
|
### Edge Detectors
|Title|
|:-:|
|[Canny edge detector](https://texonom.com/canny-edge-detector-e1ddc66952314b2f870e50621cbdbd90)|
|
7c0f6ccc9f6f49a380dbcdf347883eed
|
Image Captioning
|
Computer Vision Targets
|
Jun 7, 2023
|
Alan Jo
|
Alan Jo
|
Jun 7, 2023
|
[Text to Image](https://texonom.com/text-to-image-0dba87e7a1ef4c6ba8bc7ac0bd2c9f6f)
|
image 2 text
|
b1fa39da247648b5a514f9f4b994e4ab
|
Image Segmentation
|
Computer Vision Targets
|
Sep 14, 2020
|
Alan Jo
|
Alan Jo
|
Jun 13, 2023
|
[Data Clustering](https://texonom.com/data-clustering-cb9f2ededafa455a8757945f7c13798e)
|
## Pixel [Data Clustering](https://texonom.com/data-clustering-cb9f2ededafa455a8757945f7c13798e)
> 
> 
### Image Segmentation Notion
|Title|
|:-:|
|[Semantic Segmentation](https://texonom.com/semantic-segmentation-4b0d66cd712a4d1281feafa7988e455a)|
|[Instance Segmentation](https://texonom.com/instance-segmentation-2937d612397043d09bf7e6e60267ca34)|
|[Panoptic Segmentation](https://texonom.com/panoptic-segmentation-2dfd13d43607428f9d70f87e7b8f8f74)|
### Image Segmentation Usages
|Title|
|:-:|
|[Image Segmentation Tool](https://texonom.com/image-segmentation-tool-5fcf8df8301941e2a034027b11543ff2)|
|
eeefd926f87f457eba4034698997d86a
|
Motion Perception
|
Computer Vision Targets
|
Apr 27, 2023
|
Alan Jo
|
Alan Jo
|
Aug 5, 2023
|
### Motion Perception Notion
|Title|
|:-:|
|[Motion Field](https://texonom.com/motion-field-3d17a0f9afbc4d69abdbf5534a9fb423)|
|[Optical Flow](https://texonom.com/optical-flow-27bf0c08c5184f7dbcfbeca1afc9be5c)|
### Motion Perception Usages
|Title|
|:-:|
|[Motion Tracking](https://texonom.com/motion-tracking-9aa457c6da3442e9a087955dfdbc4edd)|
|[Motion Generating](https://texonom.com/motion-generating-2c477c23398a4cde88ef5e7b1312f41c)|
|[Motion Perception Model](https://texonom.com/motion-perception-model-a1276460fa1d43b9bbf9dcfce3611158)|
|
73bdb8160bf6470d83002d265f8306b9
|
|
Object Detection
|
Computer Vision Targets
|
Sep 22, 2020
|
Alan Jo
|
Alan Jo
|
Jun 6, 2023
|
[Feature Extraction](https://texonom.com/feature-extraction-e7c21288dbb54d218bbc25f190062fd3) [Derivative Filter](https://texonom.com/derivative-filter-1c54bd4735874bfabc93b6946b019936) [Image Segmentation](https://texonom.com/image-segmentation-eeefd926f87f457eba4034698997d86a) [AI Dataset](https://texonom.com/ai-dataset-6383f9b5b0c54e598b90dc9ea0c8e5d1)
|
## Feature Detection, Object Recognition
Template Matching is cheap when simple situation
First we have to do [Feature Extraction](https://texonom.com/feature-extraction-e7c21288dbb54d218bbc25f190062fd3)
- 2D Object Detection - Camera
- 3D Object Detection - Lidar
### Feature Detection Notion
|Title|
|:-:|
|[Bounding Box](https://texonom.com/bounding-box-34e9c2d542524787b0c2c252dfc396d6)|
|[Modeling Variability](https://texonom.com/modeling-variability-ff4ec7bdf25f46219947b4dd65f91ce1)|
|[Object Categorization](https://texonom.com/object-categorization-5d7a19e117ba4d6aa539e205cc28c1aa)|
|[Image Embedding](https://texonom.com/image-embedding-f9a37ee9964e433a9afcf05d7bf503be)|
|[Object Classification](https://texonom.com/object-classification-7000c6a471d945b191368121d0847ff6)|
|[Object Detection Dataset](https://texonom.com/object-detection-dataset-f3246358081846338249ebb7f30391eb)|
|[Texture Recognition](https://texonom.com/texture-recognition-1eb6965429da4f3d87cd98eec54b10cd)|
|[Feature Detection History](https://texonom.com/feature-detection-history-ba9d86606ca14b6aaefa2e616b3a6452)|
|[Occlusion Boundary](https://texonom.com/occlusion-boundary-0660c3c8ff14417f94e5523148bad6ff)|
### Feature Detection Usages
|Title|
|:-:|
|[Feature Detection Algorithm](https://texonom.com/feature-detection-algorithm-3190d9d22be04ef99cabce41ea172e6a)|
|[Feature Detection Object](https://texonom.com/feature-detection-object-2b32752d946d4491977af10e08c31b9c)|
|[Object Detection Model](https://texonom.com/object-detection-model-5a55a9ffae8c4e528a8081bc9d812ce8)|
> [TensorFlow.jsโ-โReal-Time Object Detection in 10 Lines of Code | Hacker Noon](https://hackernoon.com/tensorflow-js-real-time-object-detection-in-10-lines-of-code-baf15dfb95b2)
|
5f12617d24ee490ea11584ba74b110ed
|
Shape Modeling
|
Computer Vision Targets
|
Jun 28, 2021
|
Alan Jo
|
Alan Jo
|
Jun 28, 2021
|
[Pix3D](https://texonom.com/pix3d-993abbcd51c94ff9837d80d60f79ee57)
### Template Gallery
|Title|
|:-:|
|[Template Page](https://texonom.com/template-page-1834d4cbd4564fe5904e91cb0721c4de)|
|
84a770ac7869447290dfcf37ea83e465
|
|
View Synthesis
|
Computer Vision Targets
|
Mar 27, 2021
|
Alan Jo
|
Alan Jo
|
Jun 14, 2023
|
[Depth Estimation](https://texonom.com/depth-estimation-2f7e039fb9584b50b76bbb0a98bce451)
|
### View Synthesis Tools
|Title|
|:-:|
|[NeX](https://texonom.com/nex-1aeb585955ed4b59b2e0c1480a78dbd5)|
|
fb0b95a897f34fdba500a8b21aee82d0
|
Harris corner detector
|
Corner Detection Notion
|
Mar 23, 2023
|
Alan Jo
|
Alan Jo
|
Jun 6, 2023
|
[Derivative Filter](https://texonom.com/derivative-filter-1c54bd4735874bfabc93b6946b019936)
|
### Using that significant change exists to all direction in Derivative Filter at the corner
It shows orientation and magnitude
- Harris corner response is invariant to rotation ([Eigenvalue](https://texonom.com/eigenvalue-4dff6b022dd94d19b3791eef258bf9c5) is same)
- Harris corner detector is not invariant to scale
The surface $E(u,v)$ is locally approximated by a quadratic form
[Second Moment Matrix](https://texonom.com/second-moment-matrix-11b24324bf3c480eaa71412dab72369e)
## Error Function
### for each point
1. pick small region [Window Function](https://texonom.com/window-function-84df809a43bb436c8e95a78198dea06f) or [Gaussian Filter](https://texonom.com/gaussian-filter-f20f04a4f6964292a8032243414766e1)
2. compute derivatives
3. compute products of derivatives at every pixel
4. compute the sums of products of derivatives at each pixel and subtract it [DC bias](https://texonom.com/dc-bias-718d4831a85b482db2f7f7290d07bdef)
5. Compute the covariance matrix (M ๋ชจ์)
6. Compute [Eigenvector](https://texonom.com/eigenvector-d2424f7eaea4443d81105606ebf4e975) and eigenvalues
> 
7. Use threshold on a function of eigenvalues to detect corners
> 
> 
[Paraboloid](https://texonom.com/paraboloid-ea4b361db9d342909a517fdabbf50ef6)
### Pros
- Corner response R is invariant to image rotation
- R is Partial invariance to affine intensity change (shift and intensity scale)
### Limitation
- ์๋ณํ๋ง ๊ฐ์งํ๋ค
- not invariant to scale so we need multi-scale detection (canโt auto select)
- ์ฃผ๋ณ๋ง ๊ณ ๋ คํ๋ค
|
4c7d6a569a424acba2dcd39880fa3410
|
Multi-scale Blob Detection
|
Corner Detection Notion
|
Mar 28, 2023
|
Alan Jo
|
Alan Jo
|
Apr 19, 2023
|
Find local maxima in both position and scale
tradeoff
[Laplace Filter](https://texonom.com/laplace-filter-a59f4cdd29a34a669765c5cd786eb21b) we can control scale by sigma
so we can you for Gaussian pyramid
- find maximum using Harris or Laplacian
|
fbe1995acb4042d9a96608fba167afe9
|
|
Second Moment Matrix
|
Harris corner detector
| null | null | null | null | null |
|
11b24324bf3c480eaa71412dab72369e
|
Error Map
|
Depth Estimation Notion
|
Nov 10, 2021
|
Alan Jo
|
Alan Jo
|
Nov 10, 2021
|
|
5b277183ed644029adcbe0297fdc9c7d
|
|
Pinhole Camera Model
|
Depth Estimation Notion
|
Apr 11, 2023
|
Alan Jo
|
Alan Jo
|
Apr 19, 2023
|
### Single View Geometry
### Principal point
point where principal axis intersects theimage plane
- Principal Axis
> - perpendicular to the image plane
- Normalized coordinate system
> - camera center is at the origin and the principal axis is the z-axis
- f is distance between camera and image plane
- P is principal Point (0, 0, Z)
## Calibration
and then fill decomposed matrix (cameral internal and camera external becuz f is internal to the camera)
### External Factor
> 
> 
### Intrinsic Parameter
1. recover transition matrix
2. recover location
3. recover principle point and image point small x
4. use multiple and recover large X
### Two linearly independent equations
P has 11 degrees of freedom (12 parameters, but scale is arbitrary)
solve the unknown P
|
681f008c19d2456cb107e9a225a10d3a
|
|
Triangulation
|
Depth Estimation Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
### Triangulation Notion
|Title|
|:-:|
|[Linear Triangulation](https://texonom.com/linear-triangulation-30e35fd07f5d4db9a1d30be3bdbfab10)|
|[Non-Linear Triangulation](https://texonom.com/non-linear-triangulation-454567f2839c427db6dabafbe10fb6ea)|
|
086333beed5f497786fbbf8658eb4aa1
|
|
Two-view Geometry
|
Depth Estimation Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 19, 2023
|
### ํน์ง์ ์ฐพ๋๊ฑฐ ์์น ์ค์ด๊ธฐ ์ํ ๋ชฉ์ Constraint๋ก
๋ค์ค ๋ทฐ ๊ธฐํํ์์ ๋ ๊ฐ์ ์ด๋ฏธ์ง ๊ฐ์ ์๋์ ์ธ ์์น์ ๊ตฌ์กฐ๋ฅผ ๋ถ์ํ๋ ๋ฐ ์ฌ์ฉ๋๋ ๊ธฐํํ์ ๊ฐ๋
Linear Approach ํ๋ค๋ค degree of freedom so nonlinear minimize d
### Two-view Geometry Notion
|Title|
|:-:|
|[Epipolar Geomtry](https://texonom.com/epipolar-geomtry-c31dd1ad325d4a38bbad9518631df326)|
|[Binocular Stereo](https://texonom.com/binocular-stereo-00ad321ee2a64fc5894a14321fe45e0a)|
|
fe5db6a8c36e4f99abae2cb123ed1d6b
|
|
Linear Triangulation
|
Triangulation Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
30e35fd07f5d4db9a1d30be3bdbfab10
|
||
Non-Linear Triangulation
|
Triangulation Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
[Nonlinear Programming](https://texonom.com/nonlinear-programming-569f6a22fd0346a0b9ac114e849aef7d)
|
### Find X that minimizes
ํน์ง์ ์ฐพ๊ณ two-view์์ $x_1, x_2$ ๊ฐ ๊ณ์ฐ
Optimizationํ๋ ๋ฐฉ์
|
454567f2839c427db6dabafbe10fb6ea
|
Binocular Stereo
|
Two-view Geometry Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
[Image Rectification](https://texonom.com/image-rectification-7ea5336e9bc4427380bcd9d341cccfa4)
|
### Inherently fail to the flat textureless surfaces or Occlusions and repetition or Non-[Lambertian shading](https://texonom.com/lambertian-shading-aa746843f1c84f10935562169f954571) , specularities
์ฆ ํ๋ซํ๊ฑฐ๋, ๊ฐ๋ ค์ ธ ์๊ฑฐ๋ ๊ฐ๋์ ๋ฐ๋ผ ๋ณด์ด๋๊ฒ ๋ค๋ฅด๊ฑฐ๋ ๊ด์์ด ๋ฒ์ฉ์ด๊ฑฐ๋ใ
The similarity constraint is local so need to enforce global correspondenceconstraints
### Binocular Stereo Notion
|Title|
|:-:|
|[Stereo Disparity](https://texonom.com/stereo-disparity-b1f24acb19d9400bbd1fea8e07e546d5)|
|[Correspondence Search](https://texonom.com/correspondence-search-afd82937d2e34076a76a2ec76420659c)|
|[Non-local Constriant](https://texonom.com/non-local-constriant-bbc161d5e20346798fd372ffcea9fea9)|
|[Scanline stereo](https://texonom.com/scanline-stereo-2fb46aa4b34d49aebb8dd8ed0697725e)|
|[Stereo and Energy Minimization](https://texonom.com/stereo-and-energy-minimization-f49d4c56b73d4973ba4f4360efe82434)|
|[Structured Light](https://texonom.com/structured-light-687f79fb10c843e498e8223ec28e1cb1)|
|[Window Based Matching](https://texonom.com/window-based-matching-124f16fa29e14bbe9c0313e98d5d9438)|
|
00ad321ee2a64fc5894a14321fe45e0a
|
Epipolar Geomtry
|
Two-view Geometry Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
### Epipolar Geomtry Notion
|Title|
|:-:|
|[Epipolar Constraint](https://texonom.com/epipolar-constraint-8484be565211465e8de205299973623e)|
|[Fundamental Matrix](https://texonom.com/fundamental-matrix-eb229695ae624be89cc136397d621f57)|
|[Essential Matrix](https://texonom.com/essential-matrix-1af18bc29e6e485ba1b3d038a63e5876)|
|[Epipolar Plane](https://texonom.com/epipolar-plane-6f21035a8e2c48a7a229130464dbefbc)|
|[Epipole](https://texonom.com/epipole-d47a47462d774bceba39bdf6654dfa07)|
|[Two-view Baseline](https://texonom.com/two-view-baseline-569fceb255544ad793adbf67919883ca)|
|[Epipolar Line](https://texonom.com/epipolar-line-e17a20219e70489ca306a2657483cad3)|
|[The eight-point algorithm](https://texonom.com/the-eight-point-algorithm-e8dd3962126045d28072e5fde4279e53)|
|
c31dd1ad325d4a38bbad9518631df326
|
|
Correspondence Search
|
Binocular Stereo Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
afd82937d2e34076a76a2ec76420659c
|
||
Non-local Constriant
|
Binocular Stereo Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
### Global
- uniqueness
- Ordering (can be changed)
- Smoothness
|
bbc161d5e20346798fd372ffcea9fea9
|
|
Scanline stereo
|
Binocular Stereo Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
์ค์บ ๋ผ์ธ ๋จ์๋ก ์ฒ๋ฆฌํฉ๋๋ค. ๊ฐ๊ฐ์ ์ค์บ ๋ผ์ธ์ ์๋ก ๋
๋ฆฝ์ ์ผ๋ก ์ต์ ํ
๊ทธ๋์ ์ด๋ฏธ์ง ์ ์ฒด์ ๋ํ ์ผ๊ด์ฑ ์๋ ๊น์ด ์ ๋ณด๋ฅผ ์ถ์
Semi-Global Matching (SGM) ์๊ณ ๋ฆฌ์ฆ์ ๊ฒฐํฉํ์ฌ ๋ณด๋ค ์ ํํ ๊น์ด ์ ๋ณด ์ถ์ ์ ์ํ
### Shortest paths
|
2fb46aa4b34d49aebb8dd8ed0697725e
|
|
Stereo and Energy Minimization
|
Binocular Stereo Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
์๋์ง ํจ์๋ฅผ ์ต์ํํ๊ธฐ ์ํด์๋ ๊ทธ๋ํ ์ปท ์๊ณ ๋ฆฌ์ฆ์ด ์ฌ์ฉ
๊ทธ๋ํ ์์์ ์ต์ ์ปท์ ์ฐพ์๋ด๋ ์๊ณ ๋ฆฌ์ฆ์ผ๋ก, ์๋์ง ํจ์์ ์ต์๊ฐ์ ์ฐพ์๋ด๊ธฐ ์ํด ์ฌ์ฉ
์๋์ง ํจ์๋ ๋์๋๋ ํน์ง์ ์ฌ์ด์ ๊ฑฐ๋ฆฌ์ ์ผ์นํ๋ฉฐ, ๊น์ด ์ ๋ณด๋ฅผ ์ถ์ ํ๋ ๋ฐ ์ฌ์ฉ
์ผ๋ฐ์ ์ผ๋ก ๋ฐ์ดํฐ ํญ๊ณผ ์ ๊ทํ ํญ์ผ๋ก ๊ตฌ์ฑ
๋ฐ์ดํฐ ํญ์ ๋์๋๋ ํน์ง์ ์ฌ์ด์ ๊ฑฐ๋ฆฌ๋ฅผ ๋ํ๋ด๋ฉฐ, ์ ๊ทํ ํญ์ ๊น์ด ์ ๋ณด๊ฐ ์ผ๊ด์ฑ ์๋ ๊ฒฐ๊ณผ๋ฅผ ๋ด๋๋ก ์ ๋
๋ฌผ๋ฆฌํ์์ ์ฌ์ฉ๋๋ ์๋์ง ํ๋ฆ๊ณผ ์ ์ฌํ๊ธฐ๋๋ฌธ
๋ฌผ๋ฆฌํ์์์ ์๋์ง ๊ฐ๋
์ ์์คํ
๋ด๋ถ์ ์ํ๋ฅผ ๋ํ๋ด๋ฉฐ, ์์คํ
์ด ๊ฐ์ง ์๋์ง๊ฐ ์ต์ํ๋๋๋ก ์์ง์ธ๋ค๋ ๊ฒ์ ์๋ฏธ
์๋์ง ํจ์๋ ์์คํ
๋ด๋ถ์ ์ํ๋ฅผ ๋ํ๋ด๋ฉฐ, ๊น์ด ์ ๋ณด๊ฐ ์ต์ ํ๋๋๋ก ์์ง์
๋๋ค. ๋ฐ๋ผ์ ์ด๋ฅผ ์๋์ง ํจ์๋ผ๊ณ ๋ถ๋ฅธ๋ค
|
f49d4c56b73d4973ba4f4360efe82434
|
|
Stereo Disparity
|
Binocular Stereo Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 19, 2023
|
### ์ดฌ์ํ ๋ ๊ฐ์ ์ด๋ฏธ์ง์์ ๋์ํ๋ ํฝ์
๊ฐ์ ๊ฑฐ๋ฆฌ ์ฐจ์ด
depth is inversely proportional to the disparity d
baseline is $T_x$ and depth is Z
$$Z = \frac{fT_x}{d}$$
we are most interested in depth
|
b1f24acb19d9400bbd1fea8e07e546d5
|
|
Structured Light
|
Binocular Stereo Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 19, 2023
|
### Allows us to use only one camera
iPhone also use Structured Light but it is IR so canโt see
[Laser Scanning](https://texonom.com/laser-scanning-ba4f2f650af841f1bcb3bd02f03892b9)
๋ ๊ฐ์๋ฐฉ์
|
687f79fb10c843e498e8223ec28e1cb1
|
|
Window Based Matching
|
Binocular Stereo Notion
|
Apr 19, 2023
|
Alan Jo
|
Alan Jo
|
Apr 19, 2023
|
### Correspondence search
rectify the two stereo images to transformepipolar lines into scanlines
Matching cost: SSD or normalized correlation
### The similarity constraint is local so Need to enforce non-local correspondence constraints
- uniqueness
- ordering
- smoothness - disparity value change slowly (derivative)
|
124f16fa29e14bbe9c0313e98d5d9438
|
|
Laser Scanning
|
Structured Light
| null | null | null | null | null |
Project a single stripe of laser light
Scan it across the surface of the object
|
ba4f2f650af841f1bcb3bd02f03892b9
|
Epipolar Constraint
|
Epipolar Geomtry Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
ํ ์ด๋ฏธ์ง์์ ํน์ ํ ์ ์ ์ฐพ์์ ๋, ๊ทธ ์ ๊ณผ ๋์๋๋ ๋ค๋ฅธ ์ด๋ฏธ์ง ์์ ์ ์ด ์์นํ ์ ์๋ ์ ํ๋ ์์ญ์ ์ ์
์ฌ๊ธฐ์์๋ง ์์น๊ฐ๋ฅ
|
8484be565211465e8de205299973623e
|
|
Epipolar Line
|
Epipolar Geomtry Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
์ด ์์ ์ ์ด๋ผ๋ ์ ์ฝ ์กฐ๊ฑด์ ํ ์ด๋ฏธ์ง ์์ ์ ์ด ๋ค๋ฅธ ์ด๋ฏธ์ง ์์์ ์ด๋ค ์ ์์ ์์นํด์ผ ํ๋ค๋ ๊ฒ์ ์๋ฏธ
|
e17a20219e70489ca306a2657483cad3
|
|
Epipolar Plane
|
Epipolar Geomtry Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
Plane containing baseline and real target point
|
6f21035a8e2c48a7a229130464dbefbc
|
|
Epipole
|
Epipolar Geomtry Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
d47a47462d774bceba39bdf6654dfa07
|
||
Essential Matrix
|
Epipolar Geomtry Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 19, 2023
|
The essential matrix gives us the relative rotation and translation between the cameras, or their **extrinsic parameters**
|
1af18bc29e6e485ba1b3d038a63e5876
|
|
Fundamental Matrix
|
Epipolar Geomtry Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 19, 2023
|
[Essential Matrix](https://texonom.com/essential-matrix-1af18bc29e6e485ba1b3d038a63e5876)
|
Estimating the fundamental matrix is known as โweak calibrationโ
If we know the calibration matrices of the two cameras, we can estimate the essential matrix: E = KTFK
|
eb229695ae624be89cc136397d621f57
|
The eight-point algorithm
|
Epipolar Geomtry Notion
|
Apr 19, 2023
|
Alan Jo
|
Alan Jo
|
Apr 19, 2023
|
e8dd3962126045d28072e5fde4279e53
|
||
Two-view Baseline
|
Epipolar Geomtry Notion
|
Apr 13, 2023
|
Alan Jo
|
Alan Jo
|
Apr 13, 2023
|
line connecting the two camera centers
short or long
|
569fceb255544ad793adbf67919883ca
|
|
DinoV2
|
Depth Estimations
|
Apr 21, 2023
|
Alan Jo
|
Alan Jo
|
Apr 21, 2023
|
[dinov2](https://github.com/facebookresearch/dinov2)
|
12e07337f08f47ad825a1c7a51e8157b
|
|
LLFF
|
Depth Estimations
|
Mar 28, 2022
|
Alan Jo
|
Alan Jo
|
Mar 28, 2022
|
## ****Local Light Field Fusion****
[LLFF](https://github.com/Fyusion/LLFF)
|
0fb92745f0174cd59a9cb7085b335e01
|
|
Monocular Depth Estimation
|
Depth Estimations
|
Nov 10, 2021
|
Alan Jo
|
Alan Jo
|
Nov 10, 2021
|
### Monocular Depth Estimations
|Title|
|:-:|
|[
ManyDepth](https://texonom.com/manydepth-d7c5ba124f84477eb02eb7dd877bf1eb)|
|[Monodepth](https://texonom.com/monodepth-f4396afb3809474cb2fbb86acebd59b8)|
|[footprints](https://texonom.com/footprints-808a265ba81b4beabe05289b795e9147)|
|[DenseDepth](https://texonom.com/densedepth-e8c84b73b29f4abfbc59c252c94ad512)|
### Monocular Depth Estimation Tools
|Title|
|:-:|
|[Depth Hints](https://texonom.com/depth-hints-22cd35eccdf543069a4277f3237ede29)|
|[stereo-from-mono](https://texonom.com/stereo-from-mono-d3b4b226a587442fa80013b7a4a0f472)|
|
d776b0cc5eeb4288841644a89e2b7b23
|
|
NeRF
|
Depth Estimations
|
Mar 28, 2022
|
Alan Jo
|
Alan Jo
|
Jul 4, 2023
|
[nerf](https://github.com/bmild/nerf) [nerf-factory](https://github.com/kakaobrain/nerf-factory)
|
## ****Neural Radiance Fields****
****Representing Scenes as Neural Radiance Fields for View Synthesis****
### NeRFs
|Title|
|:-:|
|[Instant NeRF](https://texonom.com/instant-nerf-cd35d097f4eb49db85dcacd0c2b236eb)|
|[Gaudi AI](https://texonom.com/gaudi-ai-d87814695a144d4d9021c9721edf2717)|
|[RawNeRF](https://texonom.com/rawnerf-5b3f49c2a54447289a8e08e12b4f5c2e)|
|[ReRV](https://texonom.com/rerv-612cb54dbf034a7083913b3073593bb0)|
|[Mip-NeRF](https://texonom.com/mip-nerf-4cf012db4953490e97916d93330fd81c)|
|[Magic123](https://texonom.com/magic123-f3e7121394f242a2b41a36b74c0f7f57)|
> [๊ฐ์ง์ฐ๊ตฌ์ | Nerd's NeRF ์คํฐ๋ | D-NeRF Paper & Code Review](https://youtube.com/watch?v=y-4vLEo7hns&feature=shares)
> [NeRF: Neural Radiance Fields](https://www.matthewtancik.com/nerf)
> [Apple's new GAUDI AI turns text prompts into 3D scenes](https://mixed-news.com/en/apples-new-gaudi-ai-turns-text-prompts-into-3d-scenes)
|
199b304a6e3a4f23b1f749c0b380d03e
|
Neuralangelo
|
Depth Estimations
|
Jun 6, 2023
|
Alan Jo
|
Alan Jo
|
Jun 6, 2023
|
[NVIDIA](https://texonom.com/nvidia-42909227f1f44897ad8158cd070c0580)
|
> [Neuralangelo](https://research.nvidia.com/labs/dir/neuralangelo/)
|
0ac52e6e862f4a2d99c53d5606f57cde
|
Depth Hints
|
Monocular Depth Estimation Tools
|
Nov 10, 2021
|
Alan Jo
|
Alan Jo
|
Nov 10, 2021
|
โฃ
|
22cd35eccdf543069a4277f3237ede29
|
|
stereo-from-mono
|
Monocular Depth Estimation Tools
|
Nov 10, 2021
|
Alan Jo
|
Alan Jo
|
Nov 10, 2021
|
โฃ
|
d3b4b226a587442fa80013b7a4a0f472
|
|
DenseDepth
|
Monocular Depth Estimations
|
Nov 10, 2021
|
Alan Jo
|
Alan Jo
|
Nov 10, 2021
|
โฃ
|
e8c84b73b29f4abfbc59c252c94ad512
|
|
footprints
|
Monocular Depth Estimations
|
Nov 10, 2021
|
Alan Jo
|
Alan Jo
|
Nov 10, 2021
|
โฃ
|
๋ฐ๋ฅ ์ถ์ถ
traversable space
|
808a265ba81b4beabe05289b795e9147
|
ManyDepth
|
Monocular Depth Estimations
|
Nov 10, 2021
|
Alan Jo
|
Alan Jo
|
Nov 10, 2021
|
โฃ
|
d7c5ba124f84477eb02eb7dd877bf1eb
|
|
Monodepth
|
Monocular Depth Estimations
|
Nov 10, 2021
|
Alan Jo
|
Alan Jo
|
Nov 10, 2021
|
โฃ โฃ
|
### depth estimation from a single image
[wavelet monodepth](https://texonom.com/wavelet-monodepth-cbf90e47ff0d43ec92cd14fa42a3a4d8)
|
f4396afb3809474cb2fbb86acebd59b8
|
wavelet monodepth
|
Monodepth
| null | null | null | null | null |
โฃ
|
cbf90e47ff0d43ec92cd14fa42a3a4d8
|
Gaudi AI
|
NeRFs
|
Aug 2, 2022
|
Alan Jo
|
Alan Jo
|
Aug 2, 2022
|
[Apple Inc](https://texonom.com/apple-inc-82ae0acc12b2424bab22464a8f682e6d)
|
> [Apple's new GAUDI AI turns text prompts into 3D scenes](https://mixed-news.com/en/apples-new-gaudi-ai-turns-text-prompts-into-3d-scenes/)
|
d87814695a144d4d9021c9721edf2717
|
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.