In view of the meaning of many friends, add another,
I hope everyone will refer to the article identified by the technical information about the neural network.
My website has been served, the program is
Here
I'm too busy recently, I have no time to write. Let's take a look at it. I hope everyone will give me a suggestion. Thank you.
Graphical identification technology and online security
(Keyword: graphical identification security hacker crack OICQ login)
Author: lord
Version
right
Instant
Have
gold
sunflower
flower
work
Make
room
2004-1-4 v1.0
HOME
http://m9m.nese.net
11602011
Miracletate@sina.com
Tel
(86) 028-87311227
§1.1 directory
Graphical identification technology and online security ... 1
§1.1 directory ... 2
§1.2 Preface ... 3
§1.3 Dreams of graphical identification ... 3
§ 1.4 Graphical identification principle ... 3
§1.4.1 Graphics type ... 3
§ 1.4.1.1. Identification of regular graphics ... 3
§ 1.4.1.2. General effects graphics ... 3
§ 1.4.1.3. Identification of special graphics ... 3
§ 1.4.2 Identification principle ... 5
§ 1.4.2.1. Feature code method ... 5
§ 1.4.2.2. Handwriting flow direction method ... 5
§ 1.4.2.3. Water flow reduction method ... 6
§1.5 sample program ... 6
§1.5.1 Download address ... 6
§1.5.2 Steps ... 6
§1.2 Preface
Received:
1. This article does not have anything happen to do anything that harmful computer network! Only research discussions.
2, it is not good, I hope everyone will correct.
Today, too many websites use graphic attached code, such as
Http://love.qq.com/lex1.shtml, so that this is true, however, is this?
§1.3 Dream of graphical identification
Each programmer, each system architect, every IT person, no dream, who doesn't have a dream? I am such a person, this graphic attached identification is the first step in I have stepped out of the way of graphical identification, and I hope that the experts will do the online graphic additional code better.
§ 1.4 Principle of graphical identification
§ 1.4.1 Graphics Type
§ 1.4.1.1. Identification of regular graphics
Regular graphics refer to the graphic addition code without any special processing, such as
7564. This is the same as the width of each digit, and the height is consistent. Therefore, it is easy to identify, as long as the digital feature code of 0 to 9 is first captured, it is used as a scan comparison when it is used, and the closest is taken. Characterization collection [
Identification principle]
§ 1.4.1.2. General effect graphics
Different colors, background colors, random miscellaneous points, usually only two colors: foreground, background color. Such as:
The identification analysis principle of this graph is, first scan the entire graph, statistics, the most color: It is a background color, defined as the background color to be removed. Then you scan it again, restore the background color to white, in order to identify convenience, while removing the miscellaneous point, remove the principle of the miscellaneous point, with getpixel to remove a little bit with the background color, if different, then take the point near it, if not Continuous point, prove that it is a miscellaneous point, so use setpixel to set him. Last use
The recognition method of the conventional graphics is handled.
The sample program is to use such a method. (Sample map from:
http://verify.qq.com/getimage?",math.random (), ")
§ 1.4.1.3. Identification of special graphics
Such graphics usually have different background colors, background points, background diagrams, gradient background colors, grid backgrounds, etc., while foreground graphics (usually numbers, characters, Chinese characters, etc.) are tilted, puffed, wavy, weathering, Background graphic assembly, spherical twisted, etc. As :( The following graphic from http://edit.bjs.yahoo.com/config/eval_register?.ea=35NB3rPjpNEH191O8w9OLjUE_Nn8b16EhPzXz1BMMqq6pMb53YlmWfnAjJdcnHvar9NxW_XbmgVJsJqrxNY2ntMqZ4QY7G.1.ciCd_U.zUGUChdS9DNFevwgs3LzTTxuawVbRaoEZCggdLcEm02UCHCdaJUnb.6JSdeMyfx7Gg9eVx5mu4UH8pAiE07_yEE4doLEhWunjA--, change once every refresh)
As above, usually
The handwriting flow direction (see below) is solved, but it should be used before recognition
Waterflow reduction method restores foreground graphic from the background color grid.
§ 1.4.2 Identification principle
§ 1.4.2.1. Feature Code Method
The feature code is to count the width, height image of a single graphic word, in the statistical number of vertical, vertical image, of course, can also divide the wide level into n-equality to reduce the signature The volume of the database. Such as:
After removing excess parts, each number is 6 PIX, high 9 PIX, 0 digital graphics is: 0, 1, 2, 2, 2, 2, 2, 2, 1. That is to say: 0, 1/6, 1/3, 1/3, 1/3, 1/3, 1/3, 1/3, 1/6; vertical statistics: 4, 2, 2, 2, 4, 0, ratio: 4/9, 2/9, 2/9, 2/9, 4/9, 0. Under the C / C description
// Transverse PIX statistics (0 ~ 9)
INT NarrnumCharacterh [10] [9] = {
{0, 1, 2, 2, 2, 2, 2, 2, 1},
{0, 1, 2, 2, 1, 1, 1, 1, 5},
{0, 3, 2, 1, 1, 1, 1, 1, 5},
{0, 3, 2, 1, 2, 1, 1, 2, 3},
{0, 1, 2, 2, 2, 2, 5, 1, 1},
{0, 5, 1, 4, 2, 1, 1, 2, 3},
{0, 3, 1, 1, 4, 2, 2, 2, 3},
{0, 5, 1, 1, 1, 1, 1, 1, 1},
{0, 3, 2, 2, 3, 2, 2, 2, 3},
{0, 3, 2, 2, 2, 4, 1, 1, 3}
}
// Vertical PIX statistics (0 ~ 9)
INT NarrnumCharacters [10] [6] = {
{4, 2, 2, 2, 4, 0},
{2, 2, 8, 1, 1, 0},
{3, 3, 3, 3, 3, 0},
{2, 2, 3, 3, 5, 0},
{3, 2, 2, 8, 1, 0},
{5, 3, 3, 3, 5, 1},
{5, 3, 3, 3, 4, 0},
{1, 3, 3, 3, 2, 0},
{5, 3, 3, 3, 5, 0},
{4, 3, 3, 3, 5, 0}
}
§ 1.4.2.2. Handwriting flow
The notepad flow is to take a point when getpixel takes out a point, I want to roam out of 7 ~ 8 points around, find the right flow, if there is no surrounding
§ 1.4.2.3. Waterflow reduction method
SGFDG
§1.5 sample program
§1.5.1 Download Address
Http://m9m.nease.net/d/loadpic.exe§1.5.2 Steps
After running, click the [Acquisite Image] button, the program will automatically
http://verify.qq.com/getimage?",math.random (), "download similar
Such pictures, and saved in DL.GIF in the current directory, and then click [Analyzing] button to get the result of graphic recognition in the EDIT of [Results of the Analysis] here.
Published on March 09, 2004 10:17 AM
