Computer-vision-based framework, OpenCV

Computer Vision (CV), Department of Computer Vision (CV), Department of Computer Vision.

Graphic image Calculator Applicable on a desk, roughly three directions.

Computer Graphics, which is related to the effect and effect of the image.

Image encoder. Economical H264, H265, VP8, metropolitan image editing method.

It is a branch of the branch, a branch of the capital, a branch of the other, and a branch of the other.

Computer vision (CV).

OpenCV, a computer vision framework. No longer intel, C ++ edited for the first edition, C ++ edited for the second edition.

C language Destructor Constructor Number of Constructor Boxes, Demand and Calculators Due to computational problems, it is inconvenient to use.

Famous Mathematical Cases MATLAB, Yayoi C ++ Photographic.

C ++ Computational science area, irreplaceable status.


Lena
This is an openCV self-explanatory text, which is used in the CV area.

Lena, this is a female surname, and this is a photograph.

This is an old-fashioned department, a calculator, a calculator, and a human-like test.

After that, the number of people was developed, the calculator "Kanto-West" was effective, and the degree of emergency was very high.

At the time of the cause, the principle of deep layers of humanity (eyes, cerebrums, cerebrums, cerebrums, and credibility), complete demise.

Adequate research progress, human-like image-like image-based increase, and formation completed one-size-fits-all image-related traditional calculation method. Yu intel development, the last development industry, this opening CV.

A model of deep learning, which is a complete calculation method, a complete solution, and an incomplete solution.

Since 2016, he has been working on deep learning applications such as Goeku Yue, OpenCV Satoya added CNN type model, Can Yo Tensorflow, Caffe2.

OpenCV-like frame, mostly included

1, highGUI, this is a graphic interface, and it is effective.

Achieved highGUI, intuitional intuition and scrutiny effect.

Other common APIs imshow (), display image.

Number of images The number of functions is imshow ().

2, core, this is an opencv-like basic block, and the most basic number-setting summing method.

OpenCV-like matrix Mat, stipulated by this individual module.

Exclusion of the quadrature, and the fact that this is an individual model. Returned one, trivial, other calculation, example, meanness, difference, etc.

3, imgcodecs, pictorial editing.

Other common API Yes Two:

imread (), one image of the image, the number of images of the return result, one image, and the number of images. This is an image-like solution of the function of the function society.

imwrite (), number of images, fixed matrix, one image. For example, jpg, png formal image pieces. It is an image-like editing of the other society.

These two individual functions, re-addition imshow (), can be taken.

OpenCV unreasonable copy video piece, return possible copy video. It is possible to use a single frame, a single frame, and a single frame.

4, imgproc, image processing,

This is an OpenCV-like module, and is a multi-dimensional image processing method.

For example, Gaussian glue, Laplace transform, median wave, Gaussian wave, etc.

Gaussian Blur, Gaussian Blur, dermabrasion, which is usually used in the calculation method.


The weighted average of the weighted average image (purple in the above) and the indigo in the color of the indigo (the color of the orange), and the new image of the center of the work.

Large and small, metric, central image, metric, and homology to the above-mentioned metric. Calculation of the distribution of the two-dimensional distribution.

After the economic transition, the image was changed and smoothed, and the immediate image was changed.

Acne-like story, youth acne encounter, face, unknown face, screaming skin.

Laplace transform, Laplacian, heavy-duty image-like 边 缘 检 measurement.

， ，

Officially.

Image transformation comparatively flat region, this individual numerically small.

The image of the image is transformed into a radical region, which is a large number of individuals.

Image-like image (near the disparate area boundary line in the image), Laplace transform effect special effect. For this reason, the image segmentation is divided into Japanese images.


Image, human face, face reduction, visible, close to the eye.

Median noise, medianBlur, other action main processing noise noise.

Sichuan pepper voice, salt-and-pepper, salt-and-pepper, small quantity, salt-and-pepper.

Median wave, one area image elemental median number, action center image elemental median. A small number of noises, a medium number of impossible noises, and a number of people who have been enrolled.

Bilateral filter, bilateralFilter, this calculation complexity maximum regular wave method. At the same time, the reference image is the same as the original position difference, the Japanese color difference, and the best effect.


Right-handed filter
Left-handed photo, right-handed double-sided photograph (sweat).

The original image is on the network, and the special spots are comparatively many (laughs).

5, Local special conquest point measurement, features2d,

Local special conquest point, main image image.

After the transformation of the population into the image, it is a metamorphosis of the population special conquest.

For example, the angle of the image is different, and the angle is different.



This is an OpenCV-like self-portrait piece, a metropolitan international elephant-like chess, but a rhythmic angle.

Judgment, etc.

At this time, the inevitable transformation of the inevitable special conquest, the immediate local extreme point.

Extreme points, translations, translations, translations, and translations.

It is possible to have multiple points, and it is possible to use the same object.

Special conquest of related images, possible SIFT algorithm or SURF algorithm, OpenCV Chuto realization.

With OpenCV OpenCV2, OpenCV3, OpenCV4 Three major branches, API with minor transformation, but similar to basic reduction.

Other source code C ++ copy, provided by Ya, other language API interface, c ++-like opencv interface for bottom layer still adjustment.

"Study OpenCV 3 (Chinese version)", the latest authoritative work.

Author Nura Toshiki (Bradski), et al.

Basic books on cloth, existing OpenCV framework API usage, epipolar principles, eg epipolar orientation, epipolar orientation, etc.

Other academic OpenCV series, computer vision, computer vision, and computer vision.

This individual area has a certain degree of personality, but there is NLP.