Image description and image recognizable method

Abstract

Image description and image recognizable method, it contain (a) It obtain an image which possess plural pixels. (b) It determines a starting position in the image. (c) In the image, From the starting point along the trajectory of the former spiral aggregation makes a pixel sampling, and the pixel on the trajectory rank to the former spiral aggregation. (d) the angle increases with the increase of the variance, it forming a the angle of the latter spiral aggregation. From the starting point along a trajectory of the former spiral aggregation makes the pixel sample, and the pixel on the trajectory rank to the former spiral aggregation. (e) It decides how many frequencies the angle variation increase, and repeatedly performs the step (d). After obtaining a plurality of the latter spiral aggregation, the pixel corresponds to the value. (f) It ranks the former spiral aggregation and the latter spiral aggregation. Then, spiral aggregation map will be formed and recorded the every value of the pixel.

Claims

1. An image recognizable method for utilizing an image descriptive method to establish an image descriptive model that identify an image object with a plurality of pixels, wherein the image descriptive model with a spiral aggregation map, the method comprises: (a) determining a size of a first compared area according to a size of the image descriptive model in the image object; (b) a center point of first compared area corresponding to one of the plurality of pixels, and a trajectory of a spiral aggregation with an angle sampling the plurality of pixels in the first compared area, wherein the plurality of pixels corresponding to a value; (c) comparing the value and a value of spiral aggregation map, wherein the value of the spiral aggregation map to form a numerical distribution; (d) determining the value within the numerical distribution for recording a coordinate of the pixel correspond to the center point of first compared area; and (e) The pixel from the first compared area moving to the next pixel and executing step (d) until completely scan entire region of the image object.

2. An image recognizable method of claim 1, wherein after step (e) further comprises: (f) obtaining the recorded center point to determine a second compared area; (g) sampling the plurality of pixels located at the second compared area by a trajectory of the plurality of spiral aggregation in the second compared area, wherein the sampling pixel corresponding to a plurality of value; (h) comparing the value and a value of spiral aggregation map, wherein the spiral aggregation map includes a value and an interval number of different values; and (i) deciding the value in the interval number of different values to record for recording a coordinate of the pixel correspond to the center point.

3. An image recognizable method of claim 2, wherein after step (i) further comprises: (j) removing the recorded center point to determine a third compared area; (k) sampling the plurality of pixels located at the third compared area by a trajectory of the plurality of spiral aggregation in the third compared area to obtain a projective amount of the plurality of pixels or form a compared spiral aggregation map; (l) calculating a projective amount of the spiral aggregation map and the compared spiral aggregation map in vertical direction in the compared spiral or calculating the projective amount of the pixel and the spiral aggregation map to determine the points of the feature; (m) calculating a projective amount of the spiral aggregation map and the compared spiral aggregation map in a horizontal direction to determine a rotation of the angle; and (n) determining a relationship between the image description model and the spiral aggregation map according to the feature of the points and the rotation of the angle.

4. An image recognizable method of claim 3, wherein in step (k), the number of spiral aggregation is 360, the two adjacent spiral aggregations differ from a predetermined angle.

5. An image recognizable method of claim 1, wherein after step (g), the number of spiral aggregation is 4, and the two adjacent spiral aggregations differ from a 90 degree.

Description

BRIEF DESCRIPTION OF THE DRAWINGS

(1) FIG. 1 is the flow diagram about the image according to an embodiment of the present invention.

(2) FIG. 2 explains FIG. 1 image schematic diagram according to an embodiment of the present invention.

(3) FIG. 3 explains FIG. 1 spiral aggregation map according to an embodiment of the present invention.

(4) FIG. 4 contrast diagram about the image description method with spiral aggregation map according to an embodiment of the present invention.

(5) FIG. 5 is the flow diagram about the image description method according to an embodiment of the present invention.

(6) FIG. 6 is another flow diagram about the image description method according to an embodiment of the present invention.

(7) FIG. 7 is flow diagram about the image description method according to an embodiment of the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

(8) As indicated above, the invention provides a scenario-based security method and system. The following comprises preferred embodiments of the invention, which describe different aspects of the present invention.

(9) Referring to in FIG. 1 that is the flow diagram about the image description method of the present invention. In FIG. 1, Image descriptive method begins from step S11. As shown in FIG. 2, it obtains an image which plural pixels.

(10) Step S12, determine a starting position in the image. For example, the staring position is general on the center of the image.

(11) Step S13, in the image, from the starting point along the trajectory of the former spiral aggregation makes a pixel samples, and the pixel on the trajectory rank to the former spiral aggregation.

(12) In the specific examples, spiral aggregation definition is the dot on the spiral aggregation track. From one point has a fixed speed to the next point. It stars from the center point until the periphery finished. The spiral aggregation refers Archimedes screw geometry theorems. With the projection method formed spiral aggregation track, and it establish the image descriptive. In the characteristic of the spiral aggregation, when spiral aggregation extends outward from the center, its displacement in horizontal and vertical direction will be conducted at the same time. With the projection in the horizontal direction and vertical direction can understand the structure of the image component.

(13) Step S14, the angle increases with the increase of the variance, it forming an angle of the latter spiral aggregation. From the starting point along the trajectory of the former spiral aggregation makes a pixel sample, and the pixel on the trajectory rank to the former spiral aggregation. For example, the variance range is between 0.5 and 1 degree.

(14) Step S15, deciding how many frequency the angle variation increase, and repeatedly performs the step S14. It obtains a plurality of the latter spiral aggregation. For example, the frequent range is between 1 and 720 times.

(15) By repeating Step S14 and Step S15, the former spiral aggregation and the latter spiral aggregation rank sequentially and vertically in spiral aggregation map, as shown in FIG. 3. For example, spiral aggregation map is a rectangle. The former spiral aggregation and the latter spiral aggregation form side of the rectangles, and another side of rectangle is ranked from those pixel sampling. The detailed content will be described as blow.

(16) Step S16, arranging the former spiral aggregation and the latter spiral aggregation and form the spiral aggregation map. In other embodiment, the former spiral aggregation and the latter spiral aggregation rank sequentially and vertically in spiral aggregation map.

(17) In FIG. 3, spiral aggregation map is made from every pixel which are on different angles of spiral aggregation track. The spiral aggregation map is defined as Sam(i, ). Function indicates that the i-th pixel in the trajectory of the spiral aggregation, and spiral aggregations are on the basis of the order to rank it. In FIG. 3, horizontal direction is made from all pixels i in one spiral aggregation track. The range of pixel i is between 1 and p times. Wherein, P represents the total number of samples for the one spiral aggregation

(18) In addition, if spaced distribution rate of every spiral aggregation is angles. Thus, spiral aggregation map will get 2/ spiral aggregation in total.

(19) Referring to FIG. 4, there is the contrast diagram about the image with spiral aggregation map. In FIG. 4, the two Spiral aggregations are worth a concern. One is that the vertical line of the zero degree, another is that the vertical line of the /2 degree (the 90 degree). The spiral aggregation indicated that the horizontal direction of spiral aggregation map. If we take a vertical line in the vertical direction, it indicates that after each rotation of the spiral aggregation, the i-th pixel of the pooled. In conclusion, if the combination of all the vertical lines forms annular projective amount of the image. According to the definition of the geometry spiral theorem, there is the same spiral sampling. So, spiral aggregation will not be affected the track which is made of spiral aggregations by change the size of the image.

(20) In conclusion, spiral aggregations have two important features. First, it can resist to image size change. Because, when the size is changed, the image will maintain almost the same appearance in the vertical projection of spiral aggregations Map. Second, the rotation can be predicted. For example, after the image is rotated, will show the direction of the vertical displacement on spiral aggregations Map. With the displacement of the change quantity and direction, it can find out the differences in the direction and angle between the rotated image and the does not rotate image.

(21) Referring to FIG. 5 is the flow diagram about the image description of the present invention. In FIG. 5, image recognizable method supplies the image descriptive method to establish the image descriptive model and identify the image object. It is a plural pixel, and the image descriptive model possesses the spiral aggregations Map. The method star from step S51. Image object based on the size of the image descriptive model and decide the size in the first compared area.

(22) Step S52, the center of compared area correspond the one of the pixels, and the spiral aggregations angle of the trajectory samples the pixel in the first contrast area. The sampling pixel is corresponding to a value.

(23) Step S53, comparing the value and the value of spiral aggregation map. The value of spiral aggregation map forms a numerical distribution.

(24) Step S54, deciding whether the value falls the range of the values distribution, and recode the center point corresponding to the coordinates of the pixel.

(25) Step S55, the contrast area is moved to the next pixel, and it executes Step S54 until all the areas of the image object to complete the scan.

(26) FIG. 6 is another flow diagram about the image description of the present invention. After step S55 to step S61. By obtaining the recorded central point, it determines the second compared area.

(27) Step S62, in the second contrast area, with the trajectory of the spiral aggregations samples the pixel. The pixels is corresponding the plural values. It's worth that each spiral aggregations present sparse shape. The spiral aggregation which was selected is a representative and its image with a general description. For example, the number of spiral aggregation is 4. The two adjacent spiral aggregations differ from the 90 degree.

(28) Step S63, comparing the value and the value of the spiral aggregations. The Spiral Aggregations Map includes a value and a set of interval number of different values.

(29) Step S64, deciding the value in interval number of different values, and record that the center point is corresponding to the pixel coordinates.

(30) FIG. 7 is flow diagram about the image description of the present invention. After step S64 to step S71. By obtaining the recorded central point, it determines the third compared area. It determines the third compared area.

(31) Step S72, in the third contrast area, the intensive plurality of spiral aggregations in the trajectory samples the pixel in the contrast area. In order to receive a projective among of the pixel and form the one of spiral aggregation map. It is worth to understand that each spiral aggregations present dense shape, and the image can be completely described. For example, the number of spiral aggregation is 360. The two adjacent spiral aggregations differ from a predetermined angle. For example, this angle is 1 degree

(32) Step S73, calculating the projective amount of the vertical direction in the compared spiral aggregation map and the spiral aggregation map, and calculating the projective amount of pixel and the spiral aggregation map. It determines the points of the feature. It can decide the center position in the image.

(33) Step S74, calculating the projective amount of the horizontal direction in the compared spiral aggregation map and the spiral aggregation map. It determines the rotation of the angle.

(34) Step S75, according to the points of the feature and the angle of the rotation, it decides the relationship between the image descriptive model and spiral aggregations Map.

(35) The present invention is disclosed above by preferred embodiments. However, persons skilled in the art should understand that the preferred embodiments are illustrative of the present invention only, but should not be interpreted as restrictive of the scope of the present invention. Hence, all equivalent modifications and replacements made to the aforesaid embodiments should fall within the scope of the present invention. Accordingly, the legal protection for the present invention should be defined by the appended claims.