VAS_Image Processing

Last update:2023-10-17 18:18:09

1 VAS Intro

1.1 Brief Introduction

Integrated to CDN architecture, image processing service provides a one-stop image processing solution that includes format conversion, cropping, scaling, rotating, watermarking, etc. This service saves customers from image processing and reduces the ratio of back-to-origin requests. Besides, by outputting the adaptive content with high-reliability via the proximate CDN network, the speed of pages visiting is increased.

1.2 Applicable Product Lines

  • Content Acceleration
  • Dynamic Web Acceleration

1.3 Application Scenarios

  1. It is applicable to website customers with massive image processing demands, especially suitable for e-commerce, news, social media and UGC.
  2. It is applicable to website customers who focus on the implementation costs, performance, stability and usability of image processing.
  3. Targeting the characteristics of cellular networks and devices, image processing has better effect in mobile acceleration promotion.

Note: Some image-related services can be customized by URL rewriting. For example, there is a CDNetworks customer who was using another CDN vendor’s image optimization feature, which has parameters that are different from that of CDNetworks. And CDNetworks supports rewriting the URL by default, without changing customers usage operations. Please contact CDNetworks CSE for details.

2 VAS Detail

Image processing service includes compression, cropping, scaling, rotation, format conversion, gray scaling, watermarking, image information obtaining, etc. It supports simultaneous enabling of most operation parameters corresponding to different functions. And those parameters that do not support concurrent operations, such as file information obtaining parameters, will take effect by separate steps.

The image processing is made by an image processing server based on the parameters from the image request url. The formats of parameters from the url are fixed, which can be divided into two levels: For level-one parameters, “&” is the separator, “=” assigns the value, and the keywords are {q, Q, crop, resize, r, o, f, g, watermark, info}. For level-two parameters, “,” is the separator, “_” assigns the value, and the keywords are {p, w, h, x, y, a, g, s, e, t, r, s, image, text, d, front, color, size, bc, bs}. When a request is configured for image processing, these keywords in parameters are regarded as the image processing parameters. With the combination of multiple parameters, complex image processing can be realized to meet various user requirements of different scenarios.

2.1 Image Compression

  1. Overview
    Image quality compression. The lower the parameter of compression quality is, the higher the compression ratio and the smaller the Content-length will be.

Absolute quality

Compress the original image with q. If the original image quality is 90% and q=80, then a new image at 80% quality is generated.

Relative quality

Compress the original image with Q. If the original image quality is 90% and Q=80, then a new image at 72% quality is generated.

  1. Parameters
    Parameters: q, Q
    Value range: 1-100
    Default value: 0, no compression
  2. Example
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?q=80
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?Q=80
  3. Notices
  • When both q and Q exist, q is effective.
  • If the quality of the original image is lower than q, processing will not take effect, and the original image will be returned.
  • Suggested value: 75, as the value being too small may lead to distortion. Please be noted that the percentages used above are only for explanation, not the actual quality value of the image.
  • Only jpg and webp images formats are supported. And webp images with the parameter Q are processed by the q method by default.

2.2 Image Cropping

  1. Overview
    There are four types of image cropping: anchor point cropping, center cropping, gravity center cropping and index cropping. Besides, you can specify the cropping methods, such as square cropping, empty areas filling after cropping, etc.

Level-one parameter

Level-two parameter

Parameter value

Operation description

Value range

crop=

p_

0

Anchor point cropping

Set the anchor coordinate (x, y) in the top left corner of the original image, and crop the image with the specified width and height in pixel (w,h). When the specified height (or width) is equal to 0 or the width (or height) counting from the starting point exceeds the original image, it is directly cropped to the end of the original image. If the starting points exceed the original image, return the original image.

Example: start from coordinate x_50, y_50, crop a 400*200 image

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_0, x_50,y_50,w_400,h_200

w,h value range:[0, +∞)

1

Center cropping

Take the center of the original image as the image center, and crop out an image with the specified width and height in pixel (w,h). When either w or h is invalid, return the original image.

Example: take the original image center as the cropping center, and crop out a 400*200 image

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_1,w_400,h_200

Note: corresponding old parameters, p=4

w,h value≥0

2

Gravity center cropping

Use parameter g to set the position of nine-rectangle-grid, and crop an image with specified width and height in pixel (w,h) after offsetting x and y. When either w or h is invalid, return the original image.

Example: start from southeast, crop a 400*200 image

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_2,w_400,h_200,g_se

x,y value≥0, the default is 0;

w,h value≥0;

g value range: [nw,north,ne,west,center,east,sw,south,se].

3

Index cropping

Crop the x axis (y axis) of the original image at the specified length, and then get the image within the area with specific index. The length range is [1, length of the cropped side]. If the input length exceeds the cropped side, return the original image. The principle also applies in index cropping when the input index exceeds the maximum.

Example: crop y axis with the length of 100, and crop out the image within the corresponding area with specified index 2

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_2,i_2,y_100

x,y value≥0, default is 0;

w,h value≥0;

 i value≥0, default is 0.

4

Inscribed circle

Show the original image in a circle shape

If the final image format is png, webp or other formats that support transparency (also known as Alpha channel), then fill the non-circle area with transparency. If the final image format is jpg, fill the non-circle area with white.

Value range: r value is larger than 0 inscribed circle and scaling can be effective at the same time.

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_4,r_200

r value≥0

6

square_enlarge cropping

When the parameter s is 1, crop a square with the shorter side of the original image being the square side.

When the parameters e and s are 1, enlarge the square using the longer side of the original image.

For example: enlarge the square with the longer side of the original image being the square side.

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_6,s_1,e_1

 

s value range is0-1, no default value;

e value range is 0-1, no default value.

7

extend cropping (value of t is 1)

When the requested width and height are greater than that of the original image, center the original image and fill the surrounding with white;

When the requested width and height are smaller than that of the original image, center crop the image;

When the requested width is larger than that of the original image, and height is smaller, center crop the image and fill both left and right sides with white; When the requested width is smaller than that of the original image, and height is larger, center crop the image and fill both top and bottom sides with white;

For example: the size of original image is 640x354, and the requested size is 700x500, center crop the original image and fill the surrounding with white.

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_7,w_700,h_500

 

 

9

Percentage cropping

w, h: indicates that the width and height of thumbnail are w%W and h%H, and the value range is [1, 1000];

When w and h are specified, the image with the size of width and height in pixel (w%W, h%H) will be cropped from the start point of the original image;

When w is specified, the image with the size of width and height in pixel (w%W, w%H) will be cropped from the start point of the original image;

When h is specified, the image with the size of width and height in pixel (h%W, h%H) will be cropped from the start point of the original image.

 

x,y: are the start anchor coordinates;

When x,y are specified, the anchor coordinates are (x,y);

When x is specified, anchor coordinate is (x, (H-h)/2);

When y is specified, anchor coordinate is ((W-w)/2, y);

Note: w and h are the width and height in pixel after percentage calculation;

For example: crop an image with 50% width and 70% height of the original image.

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_9,w_50,h_70

w,h value

range is [1, 1000]

x,y value

range is [0, +∞).

 

10

Gravity center cropping after scaling

Scale down the original image, and the width and height of the re-sized image are larger or equal to w*h; Crop by specified g, and a thumbnail image of  w*h will be obtained;

When w or h is larger than that of the original image, the corresponding value of the original image is adopted;

When w or h equals to 0, make the proportional scaling with the other side;

For example: resize the original image with the proportion of 300x70, and adopt the upper part of gravity center

w,h cannot both be 0;

w,h value [0, +∞)

x,y

x-coordinate value, y-coordinate value

 

 

w,h

Width, height value

 

 

g

[nw,north,ne,west,center,east,sw,south,se]

Use parameter g to set the position of the nine-rectangle-grid

Default is se

i

Specify index block

After cropping into designated length, specify the index

Default is 0

s,e

0 or 1

Crop by a specified method

 

t

0 or 1

Crop by a specified method

 

  1. Example
    Anchor point cropping
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_0,x_50,y_50,w_400,h_200
    or
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=x_50,y_50,w_400,h_200
    Center cropping
    http://images. example.com/ctuU-fxriqqx2975432.jpg?crop=p_1,w_400,h_200
    Gravity center cropping
    http://images. example.com/ctuU-fxriqqx2975432.jpg?crop=p_2,w_400,h_200,g_se
    Index cropping
    http://images. example.com/ctuU-fxriqqx2975432.jpg?crop=p_3,i_2,y_100
    Inscribed circle
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_4,r_200
    square_enlarge cropping
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_6,s_1,e_1
    extend cropping
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_7,w_700,h_500
    Percentage cropping
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_9,w_50,h_70
    Gravity center cropping after scaling
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?crop=p_10,w_300,h_70,g_north
  2. Notices
  • Please choose the cropping method and cropping size based on the image resolution and avoid using the unreasonable parameters.
  • When the specified width (or height) from the start point exceeds that of the original image, crop the entire image of the original image.
  • To avoid potential conflicts with newly-added parameters, processing parameters of the first version image shall be fully replaced with the new ones.

2.3 Image Scaling

  1. Overview
    Image scaling is changing the width and height pixel (w for width and h for height). See below for some common scaling methods:

Level-one parameter

Level-two parameter

Parameter value

Operation description

Value range

resize=

p_

0

Compulsory stretch

Stretch the original image following w*h pixel. If there is no w*h value, use that of the original image. Parameters of w and h must coexist, otherwise the original image will be returned

Example: stretch the original image, and the width and height in pixel is 500*200

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_0,w_500,h_200

Note: corresponding old parameter p=0

w,h value range is [1-10000]

1

Percentage scaling

Resize the width and height of the original image by a percentage

For example: scale down the width to 50% of the original, and height to 50% of the original

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_1,w_50,h_50;

Note: corresponding old parameter p=5

w,h value range is [1-1000]

2

Filling after scaling

Set w and h, scale the original image proportionally. Thewidth and height should not surpass w*h. Center the re-sized image and fill the empty part with color. When either w or h is invalid, return the original image.

Example: scale down the original image to 200*200 and fill in the empty space with white

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_2,w_200,h_200;

 

w,h value range is [1-10000];

Color supports

hexadecimal, and the default is #ffffff(white)

3

Cropping after scaling

1) Specify w and h: scale down proportionally, and then crop the part that exceeds w*h, and the processed image is w*h.

2) Only specify w: if w is smaller than width and height, then scale down first before cropping, the processed image is w*w; If w is smaller than width while greater than height, then crop the image, and the processed image is w*height; If w is greater than width but smaller than height, then crop the image, and the processed image is width*w; If w is greater than width and height, the original image will be returned

3) Only specify h: if h is smaller than width and height at the same time, then scale down first before cropping, and the processed size is h*h; if h is smaller than width while greater than height, then crop the image, the processed size is h*height; if h is greater than width but smaller than height, then crop the image, and the processed size is width*h; if h is greater than width and height, return the original image

Example: scale down the original image, and the size of the processed image is 300*200

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_3,w_300,h_200;

Note: Corresponding old parameter p=1

w,h value range is [1-10000]

4

Scale down proportionally, and the width and height should not surpass w*h

Specify w and h: width is w, scale down height proportionally; height is h, scale down the width proportionally. It is correct when the width and height do not surpass w*h

2) Only specify w: w for width, scale down height proportionally

3) Only specify h: h for height, scale down width proportionally

Note: corresponding old parameter p=2

w,h value range is [1-10000]

5

Scale down proportionally, and width and height should be no smaller than w*h

1) Specify w and h: width is w, scale down width proportionally; height is h, scale down the height proportionally, and the correct result is that width and height are no smaller than w*h

2) Only specify w: when width is w, scale down height proportionally

3) Only specify h: when height is h, scale down width proportionally

Note: corresponding old parameter p=3

w,h value range is [1-10000]

6

Scale up proportionally, and if both w and h are smaller than that of the original image, width and height cannot surpass w*h

1) If width of the original image is smaller than the given w, and height of the original image is greater than the specified h, return the original image

2) If width of the original image is greater than the given w, and height of the original image is smaller than the specified h, return the original image

3) If the width of the original image is greater than the given w, and the height of the original image is greater than the specified h, scale the image according to the given w, h proportionally. Only adopt the values that do not exceed w, h.

Note: corresponding old parameter is p=8

 

w,h value range is [1-10000]

7

Proportional scaling, and the pixel should not exceed the specified value

Make the proportional scaling of the original image, and the value of width*height shall not exceed the pixel value of the specified area.

Example: specify 10000 as the pixel value, and make the proportional scaling of the image http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_7,a_10000

or:

http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=a_10000

Note: old parameters, a

a value range is [1-24999999]

 

8

Width and height limit

When the requested image is larger than a certain width and height, make the proportional scaling; the width and height of thumbnail image should be smaller or equal to the specified; or there will be no processing;

When w and h are specified, and the width and height of the original are smaller than the specified value, there will be no processing; otherwise, there will be proportional scaling down, and the width and height of thumbnail are smaller or equal to w*h;

When w is specified, and the width of the original is smaller than the specified value, there will be no processing; otherwise, there will be proportional scaling down with according to w;

When h is specified, and the height of the original is smaller than the specified value, there will be no processing; otherwise, there will be proportional scaling down according to h;

w,h value range is [1-10000]

w,h

Width, height

When value range of p is 0,2,3,4,5, w and h specify the width and height of the thumbnail

 

color

Color code

For color code, only hexadecimal value starts with # is supported

Default is #ffffff(white)

a

Width*height

When the value range of p is 7, scale the image proportionally and pixel of width*height should not exceed the a-specified pixel

A value range is [1-24999999]

  1. Example
    Compulsory stretching
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_0,w_500,h_200
    Percentage scaling
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_1,w_50,h_50
    Filling after scaling
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_2,w_200,h_200
    Cropping after scaling
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_3,w_300,h_200
    Scaling down, and the height and width should not exceed wh
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_4,w_300
    Scaling down, and the height and width should not be smaller than w
    h
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_5,w_300,h_200
    Proportional scaling, and the pixel should not exceed the specified value
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_7,a_10000
  2. Notice
    Usually image scaling down instead of scaling up is recommended in this scenario, as the oversized original image may prolong response time.
    Usually proportional scaling is more frequently adopted; width and height should not surpass w*h, and p_4 needs to be specified.
    For filling after scaling with corresponding parameters: resize=p_2,w_<width_value>,h_<hight_value>,color_<color_value>, currently it will only support jpg/jpeg format;If the width(w) or height(h) value is invalid, then the original image is returned.

2.4 Image Rotating

  1. Overview
    Rotate the image clockwise, and set the white background for rotated image.
  2. Parameters
    Parameter: r
    Value range: 0-360
    Default value: 0
  3. Example
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?r=30

2.5 Image Adaptive Rotation

  1. Overview
    Photos taken by some mobile phones may contain rotation parameters (stored in the exif information of the photo), so configuration can be set to determine whether to make adaptive rotation of the photos.
    O=0, means following the default rotation orientation of the original image instead of auto rotation;
    O=1, means auto rotating the image following the rotation parameters of the image, and if there exist the scale-down parameters, rotate the image after scaling down the image;
    O=2, means auto rotating the image following the rotation parameters of the image, and if there exist the rotation parameters, rotate the image before scaling down the image.
  2. Parameters
    Parameter: o
    Value range: [0,1,2]
    Default value: 0
  3. Example
    Scale the image proportionally, then conduct adaptive rotation
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_5,w_300&o=1
    Make adaptive rotation first, then scale the image proportionally
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?resize=p_5,w_300&o=2
  4. Notice
    Function can be realized by adopting parameter o and auto_orient that is configured in image processing, but parameter o has higher priority.

2.6 Image Format Conversion

  1. Overview
    Format conversion among four image formats (jpeg/jpg, png, gif, webp) is supported. After conversion, the respond Content-Type is the corresponding image type.
  2. Parameters
    Parameter: f
    Value range: jpg, png, gif, webp
    Default value: jpg
  3. Example
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?f=webp

2.7 Image Gray Scaling

  1. Overview
    Remove image colors and leave only gray scale.
  2. Parameters
    Parameter: g
    Value range: 0, 1
    Default value: 0
  3. Example
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?g=1
    2.8 Image Watermarking
  4. Overview
    Supports image watermarking, text watermarking and mixed watermarking

Level-one parameter

Level-two parameter

Parameters and operation description

watermark=

m_

Modes of watermark processing

0: no watermarking processing, default value;

1. Image watermarking;

2. Text watermarking;

3. Mixed watermarking, image under text;

4. Mixed watermarking, text under image;

image_

Base64 encoding, support pre-processing of watermark image

text_

Base64 encoding

d_

Watermark transparency, value range: 0-100

g_

Watermark position, value range: [nw,north,ne,west,center,east,sw,south,se]

dx_

Offset on X-axis, value range is [1,+∞, default value is 10, unit: pixel

dy_

Offset on Y-axis, value range is [1,+∞,default value is 10, unit: pixel

font_

Font of text watermark, base64 encoding

color_

Color of watermark text, with white as the default value. More options:”blue”, “#0000ff”, “rgb (0,0,255)”, base 64 encoding

size_

Font size of text watermark, value range is (0,1000], default value is 30

r_

Rotation of text watermark, value range is [0,360], default value is 0

bc_

Outlined color of the watermark text, base64 encoding

bs_

Outlined thickness of the watermark text, range value is (0, 50), default value is 1, suggested value is 0.5-2

  1. Example
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?watermark=m_1,image_aHR0cDovL2ltYWdlLWRlbW8uaW1nLWNuLWhhbmd6aG91LmFsaXl1bmNzLmNvbS9wYW5kYS5wbmc/cmVzaXplPXBfMSx3XzUwLGhfNTA=
  2. Notices
  • Parameters of image processing should be within the value range, and if the value is not specified, adopt the default values assigned.
  • Pre-processing of watermark image support percentage scaling.
  • No default value for bordercolor, and only when this field is set with a specific value, will the outlining take effect.

2.8 Image Information Obtaining

  1. Overview
    Obtain the basic or exif information of the image.
  2. Parameters
    Parameter: info
    Value range: [0,1,2], 0 means no information obtaining, 1 means basic info obtaining, and 2 means exif info obtaining.
    Default value: 0
  3. Example
    Get the basic information of the image:
    http://images.demo.cdnetworks.com//ctuU-fxriqqx2975432.jpg?info=1
    {
    “FileSize”: {“value”: “175839”},
    “Format”: {“value”: “jpeg”},
    “ImageHeight”: {“value”: “354”},
    “ImageWidth”: {“value”: “5184”},
    “XResolution”: {“value”: “72”},
    “YResolution”: {“value”: “72”},
    “ImageLength”: {“value”: “3456”},
    “BitsPerSample”: {“value”: “8, 8, 8”},
    “PhotometricInterpretation”: {“value”: “RGB”},
    “ImageDescription”: {“value”: “attends the 2016 Laureus World Sports Awards at Messe Berlin on April 18, 2016 in Berlin, Germany.”},
    “Make”: {“value”: “Canon”},
    “Model”: {“value”: “Canon EOS-1D X”},
    “SamplesPerPixel”: {“value”: “3”},
    “ResolutionUnit”: {“value”: “Inch”},
    “Software”: {“value”: “www.meitu.com”},
    “DateTime”: {“value”: “2016:04:18 19:26:05”},
    “Artist”: {“value”: “Ian Walton”},
    “YCbCrPositioning”: {“value”: “Co-sited”},
    “Copyright”: {“value”: “2016 Getty Images (Photographer) - (Editor)”},
    “ExposureTime”: {“value”: “1/250 sec.”},
    “FNumber”: {“value”: “f/7.1”},
    “ExposureProgram”: {“value”: “Manual”},
    “ISOSpeedRatings”: {“value”: “400”},
    “ExifVersion”: {“value”: “Unknown Exif Version”},
    “DateTimeOriginal”: {“value”: “2016:04:18 18:13:51”},
    “DateTimeDigitized”: {“value”: “2016:04:18 18:13:51”},
    “ComponentsConfiguration”: {“value”: “Y Cb Cr -”},
    “ShutterSpeedValue”: {“value”: “8.00 EV (1/256 sec.)”},
    “ApertureValue”: {“value”: “5.62 EV (f/7.0)”},
    “ExposureBiasValue”: {“value”: “0.00 EV”},
    “MaxApertureValue”: {“value”: “3.00 EV (f/2.8)”},
    “MeteringMode”: {“value”: “Pattern”},
    “Flash”: {“value”: “Flash did not fire, compulsory flash mode”},
    “FocalLength”: {“value”: “50.0 mm”},
    “UserComment”: {“value”: “”},
    “SubsecTime”: {“value”: “68”},
    “SubSecTimeOriginal”: {“value”: “68”},
    “SubSecTimeDigitized”: {“value”: “68”},
    “FlashPixVersion”: {“value”: “FlashPix Version 1.0”},
    “ColorSpace”: {“value”: “sRGB”},
    “PixelXDimension”: {“value”: “5184”},
    “PixelYDimension”: {“value”: “3456”},
    “FocalPlaneXResolution”: {“value”: “3545.8276”},
    “FocalPlaneYResolution”: {“value”: “3526.531”},
    “FocalPlaneResolutionUnit”: {“value”: “Inch”},
    “CustomRendered”: {“value”: “Normal process”},
    “ExposureMode”: {“value”: “Manual exposure”},
    “WhiteBalance”: {“value”: “Manual white balance”},
    “SceneCaptureType”: {“value”: “Standard”},
    “GPSVersionID”: {“value”: “2.3.0.0”}}
    If there is no EXIF information, respond message: no exif data.

2.9 Image Brightness and Contrast

  1. Overview
    Adjust image brightness and contract.
  2. Parameters
    Parameters: bright, contrast
    Value range: -100-100
    Default value: 0
  3. Example:
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?bright=10
    http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg? contrast=-10

2.10 Fuzzy Processing

  1. Overview
    Blur the image
  2. Parameter: blur

Level-two parameter

Description

Value

r

Radius of Gaussian blur

Value range is 1-50

s

standard deviation of the Gaussian distribution

Value range is 1-50

(3) Example:
http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg?blur=r_2,s_4

2.11 Background Color

  1. Overview
    Set the background color of the image.
  2. Parameter: bgcolor=
    Value range: English color words, ~hexadecimal
  3. Example:
    http://images.demo.cdnetworks.com/ctuU-fxriqqx295432.png?bgcolor=~dddfff
  4. Notice:
    Only support static image.

2.12 Interpolation

  1. Overview
    Set the interpolation algorithm of image.
  2. Parameter: interpolation=
    Value range: bilinear, bicubic, nearest-neighbor, spline
    Note: interpolation algorithm will only be effective when using with scaling.
  3. Example: http://images.demo.cdnetworks.com/ctuU-fxriqqx2975432.jpg? resize=p_0,w_100,h_100&interpolation=bilinear

3 Key Benefits

  1. Reduce storage and image optimization costs on origin sites. Customers can store only one single image, while CDNetworks’ image processing feature helps generate various images for different needs. This greatly offloads the image transformation of applications and servers at customer origin.
  2. End-users can enjoy suitable images faster from CDN edge PoPs, without any quality compromise.
Is the content of this document helpful to you?
Yes
I have suggestion
Submitted successfully! Thank you very much for your feedback, we will continue to strive to do better!