亚洲AV无码久久精品蜜桃,久久久久亚洲Av无码专,51久久夜色精品国产

DSP包含的图像算法库

polly — Tue, 18 Sep 2012 11:19:00 GMT

摘要: 阅读全文

polly 2012-09-18 19:19 发表评论

��国航空母舰�?Google Earth上的坐标

polly — Sun, 19 Aug 2012 02:34:00 GMT

Google Earth坐标�Q�美国航�I�母舰坐�?/h2>

　　�q�里�|�列�?ji��n)已�l�发现的所有美国现役和退役的航空母舰。其中包括：(x��)

　　“��鹰”�?CV63　 35°17'29.66"N,139°39'43.67"E

　　“肯尼�q?#8221;�?CVN67　 30°23'50.91"N, 81°24'14.86"W

　　“��米�?#8221;�?CVN68　 32°42'47.88"N,117°11'22.49"W

　　“艾森豪威��?#8221;�?CVN69　 36°57'27.13"N, 76°19'46.35"W

　　“林肯” �?CVN72 　 47°58'53.54"N,122°13'42.94"W

　　“华盛��?#8221;�?CVN73　 36°57'32.90"N, 76°19'45.10"W

　　“杜鲁�?#8221;�?CVN75　　36°48'53.25"N,76°17'49.29"W

　　“无畏”�?CV-11　　 40°45'53.88"N,74° 0'4.22"W

　　“莱克星顿”�?CV-2　　27°48'54.13"N,97°23'19.65"W

　　“星��”�?47°33'11.30"N,122°39'17.24"W

　　“独立”�?47°33'7.53"N,122°39'30.13"W

　　“渔R��?#8221;�?47°33'10.63"N,122°39'9.53"W

　　“�?j��ng)瑞斯�?#8221;号和“萨拉托加”受��41°31'39.59"N,71°18'58.70"W

　　“��利�?#8221;受��39°53'6.36"N,75°10'45.55"W

polly 2012-08-19 10:34 发表评论

��国��军航空母舰列表

polly — Sun, 19 Aug 2012 02:33:00 GMT

本列表收录了(ji��n)��国��军己退�Ҏ(gu��)��现役中的航空母舰�Q�包�?a class="mw-redirect" title="��国��军��C��~�号" >船��属于CV、CVA、CVB、CVL或CVN的全部舰只。编号在CVA-58之后的都属于��航空母舰�Q?a title="排水�? >排水�?/a>��过75,000吨）(j��)�Q�CVN-65和CVN-68以后的都属于核动力航�I�母�?/a>�?/p>

排水量较?y��u)��?a class="mw-redirect" title="护卫航空母舰" >护航航空母舰�Q�Escort Aircraft Carriers�Q�CVE�Q�，则另行收录于��国��军护航航空母舰列表中�?/p>

船舰�~�号	舰名		�U�别	附注
CV-1	Langley	兰利�?/a>		以运煤舰朱比特号�Q�USS Jupiter�Q�改造而成
CV-2	Lexington	列克星敦�?/a>	5�?�?/a>CV-3	Saratoga	萨拉托加�?/a>	列克星敦�U?/td>	7�?5�?/a>�?a title="比基��环�C? >比基��环�C?/a>的核子武器试验中沉没
CV-4	Ranger	�H�击者号	�H�击者��	10�?8�?/a>退�?/td>
CV-5	Yorktown	�U�克城号	�U�克城��	6�?�?/a>�?a title="中途岛��h��" >中途岛��h��中沉�?/td>
CV-6	Enterprise	企业�?/a>	�U�克城��	2�?7�?/a>退�?/td>
CV-7	Wasp	胡蜂�?/a>	9�?5�?/a>�?a class="mw-redirect" title="日军" >日军潜艇�?y��n)L��
CV-8	Hornet	大黄蜂号	�U�克城��	10�?7�?/a>�?a title="圣克鲁斯��岛战役" >圣克鲁斯��岛战役中受重创沉没
CV-9	Essex	6�?0�?/a>退�?/td>
CV-10	Yorktown	�U�克城号	埃塞克斯�U?/td>	6�?7�?/a>退�?/td>
CV-11	Intrepid	埃塞克斯�U?/td>	3�?5�?/a>退�?/td>
CV-12	Hornet	大黄蜂号	埃塞克斯�U?/td>	6�?4�?/a>退�?/td>
CV-13	Franklin	埃塞克斯�U?/td>	2�?7�?/a>退�?/td>
CV-14	Ticonderoga	提康��L(f��ng)��加号	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CV-15	Randolph	伦道夫号	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CV-16	Lexington	列克星敦�?/a>	埃塞克斯�U?/td>	11�?�?/a>退�?/td>
CV-17	Bunker Hill	邦克山号	埃塞克斯�U?/td>	1�?�?/a>退�?/td>
CV-18	Wasp	胡蜂�?/a>	埃塞克斯�U?/td>	7�?�?/a>退�?/td>
CV-19	Hancock	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CV-20	Bennington	本宁��号	埃塞克斯�U?/td>	1�?5�?/a>退�?/td>
CV-21	Boxer	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CVL-22	Independence	独立�?/a>	�?#8220;克里夫兰�U�轻巡洋�?#8221;改装而成
CVL-23	Princeton	普林斯顿�?/a>	独立�U?/td>	�?#8220;克里夫兰�U�轻巡洋�?#8221;改装而成
CVL-24	Belleau Wood	独立�U?/td>	�?#8220;克里夫兰�U�轻巡洋�?#8221;改装而成
CVL-25	Cowpens	�U�本斯号	独立�U?/td>	�?#8220;克里夫兰�U�轻巡洋�?#8221;改装而成
CVL-26	Monterey	蒙特利号	独立�U?/td>	�?#8220;克里夫兰�U�轻巡洋�?#8221;改装而成
CVL-27	Langley	兰利�?/a>	独立�U?/td>	�?#8220;克里夫兰�U�轻巡洋�?#8221;改装而成
CVL-28	Cabot	卡伯特号	独立�U?/td>	�?#8220;克里夫兰�U�轻巡洋�?#8221;改装而成
CVL-29	Bataan	独立�U?/td>	�?#8220;克里夫兰�U�轻巡洋�?#8221;改装而成
CVL-30	San Jacinto	独立�U?/td>	�?#8220;克里夫兰�U�轻巡洋�?#8221;改装而成
CV-31	Bon Homme Richard	好�h理查德号	埃塞克斯�U?/td>	7�?�?/a>退�?/td>
CV-32	Leyte	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CV-33	Kearsarge	奇沙��d��	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CV-34	Oriskany	奥里斯卡��号	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CV-35	Reprisal	埃塞克斯�U?/td>	建造中途取�?/td>
CV-36	Antietam	安提坦号	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CV-37	Princeton	普林斯顿�?/a>	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CV-38	Shangri-la	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CV-39	Lake Champlain	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CV-40	Tarawa	塔拉瓦号	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CVB-41	Midway	中途岛�U?/a>	4�?1�?/a>退�?/td>
CVB-42	Franklin D. Roosevelt	中途岛�U?/td>
CVB-43	Coral Sea	中途岛�U?/td>
CVB-44		�?/td>		建造计划取�?/td>
CV-45	Valley Forge	��吉谷号	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CV-46	Iwo Jima	��磺岛号	埃塞克斯�U?/td>	建造计划取�?/td>
CV-47	Philippine Sea	埃塞克斯�U?/td>	长舰体埃塞克斯��Q�Long-hull Essex�Q?/td>
CVL-48	Saipan	1�?4�?/a> 正式除役
CVL-49	Wright	��q��?/a>	塞班岛��	5�?7�?/a> 正式除役
CV-50到CV-55		�?/td>	埃塞克斯�U?/td>	建造计划取�?/td>
CVB-56到CVB-57		�?/td>	中途岛�U?/td>	建造中途取�?/td>
CVA-58	United States	建造中途取�?/td>
CVA-59	Forrestal	��莱斯特�U?/a>	9�?1�?/a> 正式除役
CVA-60	Saratoga	��莱斯特�U?/td>	8�?0�?/a> 正式除役
CVA-61	Ranger	��莱斯特�U?/td>	7�?0�?/a> 正式除役
CV-62	Independence	��莱斯特�U?/td>	9�?0�?/a> 正式除役
CV-63	Kitty Hawk	5�?2�?/a> 正式除役
CV-64	Constellation	��鹰�U?/td>	8�?�?/a> 正式除役
CVN-65	Enterprise	企业�?/a>	企业�U?/td>	服役�?/td>
CVA-66	America	��鹰�U?/td>	8�?�?/a> 正式除役
CV-67	John F. Kennedy	�Q�改良）(j��)��鹰�U?/td>	8�?�?/a> 正式除役
CVN-68	Nimitz	��米兹号	服役�?/td>
CVN-69	Dwight D. Eisenhower	��米兹��	服役�?/td>
CVN-70	Carl Vinson	��米兹��	服役�?/td>
CVN-71	Theodore Roosevelt	�\|�斯��号	��米兹��	服役�?/td>
CVN-72	Abraham Lincoln	��米兹��	服役�?/td>
CVN-73	George Washington	华盛��号	��米兹��	服役�?/td>
CVN-74	John C. Stennis	��米兹��	服役�?/td>
CVN-75	Harry S. Truman	杜鲁门号	��米兹��	服役�?/td>
CVN-76	Ronald Reagan	��米兹��	服役�?/td>
CVN-77	George H. W. Bush	��米兹��	服役�?/td>
CVN-78	Gerald R. Ford	建造中
CVN-79	John F. Kennedy	肯尼�q�号	��特�U?/td>	建造中
CVN-80		未命�?/a>	��特�U?/td>	计划�?/td>

polly 2012-08-19 10:33 发表评论

polly — Fri, 10 Aug 2012 02:42:00 GMT

高光谱成像是��C��代光甉|��技术，兴�v�?O世纪8O�q�代�Q�目前仍在迅猛发展巾。高光谱成像是相对多光谱成像而言�Q�通过高光谱成像方法获得的高光谱图像与通过多光谱成像获取的多光谱图像相比具有更丰富的图像和光谱信息。如果根据传感器的光谱分辨率对光谱成像技术进行分�c�，光谱成像技术一般可分成3�c�R�?/p>

(1) 多光谱成�?#8212;—光谱分��L率在 delta_lambda/lambda=0�Q?数量�U�，�q�样的传感器在可见光和近�U�外区域一般只有几个�L�D�c(di��n)�?/p>

(2) 高光谱成�?#8212;— 光谱分��L率在 delta_lambda/lambda=0�Q?1数量�U�，�q�样的传感器在可见光和近�U�外区域有几卜到数百个�L�D�，光谱分��L率可达nm�U��?/p>

(3) ��光谱成�?#8212;— 光谱分��L率在delta_lambda/lambda =O�Q?01数量�U�，�q�样的传感器在可见光和近�U�外区域可达数千个�L�D�c(di��n)�?/p>

众所周知�Q�光谱分析是自然�U�学中一�U�重要的研究手段�Q�光谱技术能��(g��)��到被测物体的物理结构、化学成分等指标。光��p��h��Z��Ҏ(gu��)��量，而图像测量是��Z��I�间�Ҏ(gu��)��变化，两者各有其优缺炏V��因此，可以说光谱成像技术是光谱分析技术和囑փ�分析技术发展的必然�l�果�Q�是二者完��结合的产物。光谱成像技术不仅具有光谱分辨能力，�q�具有图像分辨能力，利用光谱成像技术不仅可以对待检��物体进行定性和定量分析�Q�而且�q�能�q�对其进行定位分析�?/p>

高光谱成像系�l�的主要工作部�g是成像光�׃?n��i)A�Q�它是一�U�新型传感器�Q?O世纪8O�q�代初正式开始研�Ӟ��研制�q�类仪器的目的是��取大量窄(ji��ng)波段�q�箋(hu��)光谱囑փ�数据�Q��每个像元��h��几乎�q�箋(hu��)的光谱数据。它是一�p�d��光�L波长处的光学囑փ��Q�通常包含数十到数百个波段�Q�光谱分辨率一般�ؓ(f��)1～l0nm。由于高光谱成像所获得的高光谱囑փ�能对囑փ�中的每个像素提供一条几乎连�l�的光谱曲线�Q�其在待��物上获得空间信息的同时又能获得比多光谱更�ؓ(f��)丰富光谱数据信息�Q�这些数据信息可用来生成复杂模型�Q�来�q�行判别、分�c�R��识别图像中的材料�?/p>

通过高光谱成像获取待��物的高光谱囑փ�包含�?ji��n)待��物的丰富的�I�间、光谱和辐射三重信息。这些信息不仅表��C��(ji��n)

地物�I�间分布的媄(ji��ng)像特征，同时也可能以其中某一像元或像元组为目标获取它们的辐射强度以及(qi��ng)光谱特征。媄(ji��ng)像、辐��与光谱是高光谱囑փ�中的3个重要特征，�q?个特征的有机�l�合��是高光谱图像�?/p>

高光谱图像数据�ؓ(f��)数据立方�?cube)。通常囑փ�像素的横坐标和纵坐标分别用z和Y来表�C�，光谱的�L长信息以(Z卌��u)表示。该数据立方体由沿着光谱轴的以一定光谱分辨率间隔的连�l�二�l�图像组成�?/p>

polly 2012-08-10 10:42 发表评论

polly — Wed, 25 Jul 2012 11:02:00 GMT

��法效率�Q�先验特征，��法框架本周搞定�?img src ="http://www.shnenglu.com/polly-yang/aggbug/185048.html" width = "1" height = "1" />

polly 2012-07-25 19:02 发表评论

Bilateral Filtering for Gray and Color Images

polly — Tue, 24 Jul 2012 12:39:00 GMT

Introduction
The Idea
The Gaussian Case
Experiments with Black-and-White Images
Experiments with Color Images
References

Introduction

Filtering is perhaps the most fundamental operation of image processing and computer vision. In the broadest sense of the term "filtering", the value of the filtered image at a given location is a function of the values of the input image in a small neighborhood of the same location. For example, Gaussian low-pass filtering computes a weighted average of pixel values in the neighborhood, in which the weights decrease with distance from the neighborhood center. Although formal and quantitative explanations of this weight fall-off can be given, the intuition is that images typically vary slowly over space, so near pixels are likely to have similar values, and it is therefore appropriate to average them together. The noise values that corrupt these nearby pixels are mutually less correlated than the signal values, so noise is averaged away while signal is preserved.
The assumption of slow spatial variations fails at edges, which are consequently blurred by linear low-pass filtering. How can we prevent averaging across edges, while still averaging within smooth regions? Many efforts have been devoted to reducing this undesired effect. Bilateral filtering is a simple, non-iterative scheme for edge-preserving smoothing.

Back to Index

The Idea

The basic idea underlying bilateral filtering is to do in the range of an image what traditional filters do in its domain. Two pixels can be close to one another, that is, occupy nearby spatial location, or they can be similar to one another, that is, have nearby values, possibly in a perceptually meaningful fashion.
Consider a shift-invariant low-pass domain filter applied to an image:

The bold font for f and h emphasizes the fact that both input and output images may be multi-band. In order to preserve the DC component, it must be

Range filtering is similarly defined:

In this case, the kernel measures the photometric similarity between pixels. The normalization constant in this case is

The spatial distribution of image intensities plays no role in range filtering taken by itself. Combining intensities from the entire image, however, makes little sense, since the distribution of image values far away from x ought not to affect the final value at x. In addition, one can show that range filtering without domain filtering merely changes the color map of an image, and is therefore of little use. The appropriate solution is to combine domain and range filtering, thereby enforcing both geometric and photometric locality. Combined filtering can be described as follows:

with the normalization

Combined domain and range filtering will be denoted as bilateral filtering. It replaces the pixel value at x with an average of similar and nearby pixel values. In smooth regions, pixel values in a small neighborhood are similar to each other, and the bilateral filter acts essentially as a standard domain filter, averaging away the small, weakly correlated differences between pixel values caused by noise. Consider now a sharp boundary between a dark and a bright region, as in figure 1(a).


(a)	(b)	(c)
	Figure 1

When the bilateral filter is centered, say, on a pixel on the bright side of the boundary, the similarity function s assumes values close to one for pixels on the same side, and values close to zero for pixels on the dark side. The similarity function is shown in figure 1(b) for a 23x23 filter support centered two pixels to the right of the step in figure 1(a). The normalization term k(x) ensures that the weights for all the pixels add up to one. As a result, the filter replaces the bright pixel at the center by an average of the bright pixels in its vicinity, and essentially ignores the dark pixels. Conversely, when the filter is centered on a dark pixel, the bright pixels are ignored instead. Thus, as shown in figure 1(c), good filtering behavior is achieved at the boundaries, thanks to the domain component of the filter, and crisp edges are preserved at the same time, thanks to the range component.

Back to Index

The Gaussian Case

A simple and important case of bilateral filtering is shift-invariant Gaussian filtering, in which both the closeness function c and the similarity function s are Gaussian functions of the Euclidean distance between their arguments. More specifically, c is radially symmetric:

where

is the Euclidean distance. The similarity function s is perfectly analogous to c :

where

is a suitable measure of distance in intensity space. In the scalar case, this may be simply the absolute difference of the pixel difference or, since noise increases with image intensity, an intensity-dependent version of it. Just as this form of domain filtering is shift-invariant, the Gaussian range filter introduced above is insensitive to overall additive changes of image intensity. Of course, the range filter is shift-invariant as well.

Back to Index

Experiments with Black-and-White Images

Figure 2 (a) and (b) show the potential of bilateral filtering for the removal of texture. The picture "simplification" illustrated by figure 2 (b) can be useful for data reduction without loss of overall shape features in applications such as image transmission, picture editing and manipulation, image description for retrieval.


(a)		(b)
	Figure 2

Bilateral filtering with parameters sd =3 pixels and sr =50 intensity values is applied to the image in figure 3 (a) to yield the image in figure 3 (b). Notice that most of the fine texture has been filtered away, and yet all contours are as crisp as in the original image. Figure 3 (c) shows a detail of figure 3 (a), and figure 3 (d) shows the corresponding filtered version. The two onions have assumed a graphics-like appearance, and the fine texture has gone. However, the overall shading is preserved, because it is well within the band of the domain filter and is almost unaffected by the range filter. Also, the boundaries of the onions are preserved.


(a)		(b)

(c)		(d)
	Figure 3

Back to Index

Experiments with Color Images

For black-and-white images, intensities between any two gray levels are still gray levels. As a consequence, when smoothing black-and-white images with a standard low-pass filter, intermediate levels of gray are produced across edges, thereby producing blurred images. With color images, an additional complication arises from the fact that between any two colors there are other, often rather different colors. For instance, between blue and red there are various shades of pink and purple. Thus, disturbing color bands may be produced when smoothing across color edges. The smoothed image does not just look blurred, it also exhibits odd-looking, colored auras around objects.


(a)		(b)

(c)		(d)
	Figure 4

Figure 4 (a) shows a detail from a picture with a red jacket against a blue sky. Even in this unblurred picture, a thin pink-purple line is visible, and is caused by a combination of lens blurring and pixel averaging. In fact, pixels along the boundary, when projected back into the scene, intersect both red jacket and blue sky, and the resulting color is the pink average of red and blue. When smoothing, this effect is emphasized, as the broad, blurred pink-purple area in figure 4 (b) shows.
To address this difficulty, edge-preserving smoothing could be applied to the red, green, and blue components of the image separately. However, the intensity profiles across the edge in the three color bands are in general different. Smoothing the three color bands separately results in an even more pronounced pink and purple band than in the original, as shown in figure 4 (c). The pink-purple band, however, is not widened as in the standard-blurred version of figure 4 (b).
A much better result can be obtained with bilateral filtering. In fact, a bilateral filter allows combining the three color bands appropriately, and measuring photometric distances between pixels in the combined space. Moreover, this combined distance can be made to correspond closely to perceived dissimilarity by using Euclidean distance in the CIE-Lab color space. This color space is based on a large body of psychophysical data concerning color-matching experiments performed by human observers. In this space, small Euclidean distances are designed to correlate strongly with the perception of color discrepancy as experienced by an "average" color-normal human observer. Thus, in a sense, bilateral filtering performed in the CIE-Lab color space is the most natural type of filtering for color images: only perceptually similar colors are averaged together, and only perceptually important edges are preserved. Figure 4 (d) shows the image resulting from bilateral smoothing of the image in figure 4 (a). The pink band has shrunk considerably, and no extraneous colors appear.


(a)	(b)	(c)
	Figure 5

Figure 5 (c) shows the result of five iterations of bilateral filtering of the image in figure 5 (a). While a single iteration produces a much cleaner image (figure 5 (b)) than the original, and is probably sufficient for most image processing needs, multiple iterations have the effect of flattening the colors in an image considerably, but without blurring edges. The resulting image has a much smaller color map, and the effects of bilateral filtering are easier to see when displayed on a printed page. Notice the cartoon-like appearance of figure 5 (c). All shadows and edges are preserved, but most of the shading is gone, and no "new" colors are introduced by filtering.

Back to Index

References

[1] C. Tomasi and R. Manduchi, "Bilateral Filtering for Gray and Color Images", Proceedings of the 1998 IEEE International Conference on Computer Vision, Bombay, India.
[2] T. Boult, R.A. Melter, F. Skorina, and I. Stojmenovic,"G-neighbors", Proceedings of the SPIE Conference on Vision Geometry II, pages 96-109, 1993.
[3] R.T. Chin and C.L. Yeh, "Quantitative evaluation of some edge-preserving noise-smoothing techniques", Computer Vision, Graphics, and Image Processing, 23:67-91, 1983.
[4] L.S. Davis and A. Rosenfeld, "Noise cleaning by iterated local averaging", IEEE Transactions on Systems, Man, and Cybernetics, 8:705-710, 1978.
[5] R.E. Graham, "Snow-removal - a noise-stripping process for picture signals", IRE Transactions on Information Theory, 8:129-144, 1961.
[6] N. Himayat and S.A. Kassam, "Approximate performance analysis of edge preserving filters", IEEE Transactions on Signal Processing, 41(9):2764-77, 1993.
[7] T.S. Huang, G.J. Yang, and G.Y. Tang, "A fast two-dimensional median filtering algorithm", IEEE Transactions on Acoustics, Speech, and Signal Processing, 27(1):13-18, 1979.
[8] J.S. Lee, "Digital image enhancement and noise filtering by use of local statistics", IEEE Transactions on Pattern Analysis and Machine Intelligence, 2(2):165-168, 1980.
[9] M. Nagao and T. Matsuyama, "Edge preserving smoothing", Computer Graphics and Image Processing, 9:394-407, 1979.
[10] P.M. Narendra, "A separable median filter for image noise smoothing", IEEE Transactions on Pattern Analysis and Machine Intelligence, 3(1):20-29, 1981.
[11] K.J. Overton and T.E. Weymouth, "A noise reducing preprocessing algorithm",Proceedings of the IEEE Computer Science Conference on Pattern Recognition and Image Processing, pages 498-507, Chicago, IL, 1979.
[12] P. Perona and J. Malik, "Scale-space and edge detection using anisotropic diffusion", IEEE Transactions on Pattern Analysis and Machine Intelligence, 12(7):629-639, 1990.
[13] G. Ramponi, "A rational edge-preserving smoother", Proceedings of the International Conference on Image Processing, volume 1, pages 151-154, Washington, DC, 1995.
[14] G. Sapiro and D.L. Ringach, "Anisotropic diffusion of color images", Proceedings of the SPIE, volume 2657, pages 471-382, 1996.
[15] D.C.C. Wang, A.H. Vagnucci, and C.C. Li, "A gradient inverse weighted smoothing scheme and the evaluation of its performance", Computer Vision, Graphics, and Image Processing, 15:167-181, 1981.
[16] G. Wyszecki and W. S. Styles, Color Science: Concepts and Methods, Quantitative Data and Formulae, John Wiley and Sons, New York, NY, 1982.
[17] L. Yin, R. Yang, M. Gabbouj, and Y. Neuvo, "Weighted median filters: a tutorial",IEEE Transactions on Circuits and Systems II: Analog and Digital Signal Processing, 43(3):155-192, 1996.

polly 2012-07-24 20:39 发表评论

VS2010中配�|�opencv和Python2.6

polly — Sat, 21 Jul 2012 07:54:00 GMT

前提�Q?我自己已�l�安装好VS2010 + opencv2.4
1.下蝲python 2.6
2.下蝲numpy�Q�此处需要下载对应python版本的，如numpy-1.6.1-win32-superpack-python2.6.exe

3.然后��此目录下的cv2.pyd文�g�Q�即D:\opencv2.4\build\python\2.6\cv2.pyd文�g�Q�复制到你的python�?span style="color: #cc0000">site-packages目录下面�Q�比如我的是C:\Python26\Lib\site-packages�q�个目录下面.

好了(ji��n)�Q�测试一下：(x��)

�q�入D:\opencv2.4\samples\python双击drawing.py或者cd到目录下面然后python drawing.py

polly 2012-07-21 15:54 发表评论

Gabor滤�L��结

polly — Sat, 14 Jul 2012 03:09:00 GMT

摘要: 转蝲��h��明：(x��)http://www.shnenglu.com/polly-yang/一�Q�房屋检��小�l?nbsp; 一开始，直接用LSD�Q�Line Segment Detector�Q�检��VHR�Q�Very High Resolution�Q�遥感卫星图像中的房屋，效果很屎。效果很屎的主要原因是因为存在各�U�干扎ͼ�概括下来�Q�主要有�Q?nbs... 阅读全文

polly 2012-07-14 11:09 发表评论

Mark几篇文章

polly — Wed, 11 Jul 2012 09:08:00 GMT

2011 , A Probabilistic Framework to Detect Building in Aerial and Satellite Images.
2011 , Saliency and Gist Features for Target Detection in Satellite Images.
回头再�ȝ��

polly 2012-07-11 17:08 发表评论

polly — Wed, 08 Feb 2012 08:34:00 GMT

��q��矩阵用两个位�|�的象素的联合概率密度来定义�Q�它不仅反映亮度的分布特性，也反映具有同样亮度或接近亮度的象素之间的位置分布�Ҏ(gu��)��，是有兛_��象亮度变化的二阶�l�计特征。它是定义一�l�纹理特征的基础�?

一�q�图象的灰度��q��矩阵能反映出图象灰度关于方向、相邻间隔、变化幅度的�l�合信息�Q�它是分析图象的局部模式和它们排列规则的基��?/p>

　　设f(x,y)��Z��q�二�l�数字图象，其大��ؓ(f��)M×N�Q�灰度��别�ؓ(f��)Ng,则满��一定空间关�pȝ��灰度��q��矩阵�?/p>

P(i,j)=#�?x1,y1),(x2,y2)∈M×N｜f(x1,y1)=i,f(x2,y2)=j�?/p>

　　其中#(x)表示集合x中的元素个数�Q�显然P为Ng×Ng的矩阵，�?x1,y1)�?x2,y2)间距��Mؓ(f��)d,两者与坐标横��u的夹角�ؓ(f��)θ�Q�则可以得到各种间距�?qi��ng)角度的灰度��q��矩阵P(i,j,d,θ)�?/p>

�U�理特征提取的一�U�有效方法是以灰度��的空间相关矩阵即��q��矩阵为基��的��E7�Q�，因�ؓ(f��)囑փ�中相�?Δx�Q?#916;y)的两个灰度像素同时出现的联合频率分布可以用灰度共生矩阉|��表示。若��图像的灰度�U�定为N�U�，那么��q��矩阵为N×N矩阵�Q�可表示为M(Δx�Q?#916;y)(h,k)�Q�其中位�?h,k)的元素mhk的��D��C�Z��个灰度�ؓ(f��)h而另一个灰度�ؓ(f��)k的两个相距�ؓ(f��)(Δx�Q?#916;y)的像素对出现的次数�?br />　　对粗�U�理的区域，其灰度共生矩�늚�mhk��D��集中于主对角�U�K��q�。因为对于粗�U�理�Q�像素对��于��h��相同的灰度。而对于细�U�理的区域，其灰度共生矩阵中的mhk值则散布在各处�?/p>

��Z��(ji��n)能更直观��C��q��矩阵描述�U�理状况�Q�从��q��矩阵导出一些反映矩�늊��늚�参数�Q�典型的有以下几�U�：(x��)

�Q?�Q�能量：(x��) 是灰度共生矩阵元素值的�q�x(ch��ng)��和，所以也�U�能量，反映�?ji��n)图像灰度分布均匀�E�度和纹理粗�l�度。如果共生矩�늚�所有值均相等�Q�则ASM值小�Q�相反，如果其中一些值大而其它值小�Q�则ASM值大。当��q��矩阵中元素集中分布时�Q�此时ASM值大。ASM值大表明一�U�较均一和规则变化的�U�理模式�?/p>

�Q?�Q�对比度�Q?�Q�其�?。反映了(ji��n)囑փ�的清晰度和纹理沟�UҎ(gu��)��的�E�度。纹理沟�U�越深，其对比度��大�Q�视觉效果越清晰�Q�反之，�Ҏ(gu��)��度小�Q�则沟纹��，效果模糊。灰度差卛_��比度大的象素对越多，�q�个��D��大。灰度公生矩阵中�q�离对角�U�的元素��D��大，CON��大�?/p>

�Q?�Q�相养I��(x��)它度量空间灰度共生矩阵元素在行或列方向上的相似程度，因此�Q�相兛_��大��反映了(ji��n)囑փ�中局部灰度相��x(ch��ng)��。当矩阵元素值均匀相等�Ӟ��相关值就�?相反�Q�如果矩阵像元值相差很大则相关值小。如果图像中有水�q�x(ch��ng)��向纹理，则水�q�x(ch��ng)��向矩�늚�COR大于其余矩阵的COR倹{�?/p>

�Q?�Q�熵�Q?是图像所��h��的信息量的度量，�U�理信息也属于图像的信息�Q�是一个随机性的度量�Q�当��q��矩阵中所有元素有最大的随机性、空间共生矩阵中所有值几乎相�{�时�Q�共生矩阵中元素分散分布�Ӟ��熵较大。它表示�?ji��n)图像中�U�理的非均匀�E�度或复杂程度�?/p>

�Q?�Q�逆差距：(x��) 反映囑փ��U�理的同质性，度量囑փ��U�理局部变化的多少。其值大则说明图像纹理的不同区域间缺��变化，局部非常均匀�?br />

其它参数:

中�?lt;Mean>

协方�?lt;Variance>

同质�?逆差�?lt;Homogeneity>

反差

差异�?lt;Dissimilarity>

�?lt;Entropy>

二阶�?lt;Angular Second Moment>

自相�?lt;Correlation>

polly 2012-02-08 16:34 发表评论

polly — Thu, 02 Feb 2012 05:24:00 GMT

PCA-Principal Components Analysis
下面我就对PCA做一个简单的介绍吧：(x��)

PCA是主成分分析�Q�主要用于数据降�l�_(d��)��对于一�p�d��sample的feature�l�成的多�l�向量，多维向量里的某些元素本��n没有区分性，比如某个元素在所有的sample中都�?�Q�或者与1差距不大�Q�那么这个元素本�w�就没有区分性，用它做特征来区分�Q��A(ch��)献会(x��)非常��。所以我们的目的是找那些变化大的元素�Q�即方差大的那些�l�_(d��)��而去除掉那些变化不大的维�Q�从而��feature留下的都�?#8220;�_�֓�”�Q�而且计算量也变小�?ji��n)�?/p>

对于一个k�l�的feature来说�Q�相当于它的每一�l�feature与其他维都是正交的（相当于在多维坐标�p�M��Q�坐标��u都是垂直的）(j��)�Q�那么我们可以变化这些维的坐标系�Q�从而�ɘq�个feature在某些维上方差大�Q�而在某些�l�上方差很小。例如，一�?5度倾斜的椭圆，在第一坐标�p�，如果按照x,y坐标来投影，�q�些点的x和y的属性很隄��于区分他们，因�ؓ(f��)他们在x,y轴上坐标变化的方差都差不多，我们无法�Ҏ(gu��)��q�个点的某个x属性来判断�q�个�Ҏ(gu��)��哪个�Q�而如果将坐标轴旋转，以椭圆长轴�ؓ(f��)x��_(d��)��则椭圆在长��u上的分布比较长，方差大，而在短��u上的分布短，方差��，所以可以考虑只保留这些点的长轴属性，来区分椭圆上的点�Q�这��P��区分性比x,y轴的�Ҏ(gu��)��要好�Q?/p>

所以我们的做法��是求得一个k�l�特征的投媄(ji��ng)矩阵�Q�这个投��q��阵可以将feature从高�l�降��C��l�。投��q��阵也可以叫做变换矩阵。新的低�l�特征必��L��个维都正交，特征向量都是正交的。通过求样本矩�늚�协方差矩阵，然后求出协方差矩�늚�特征向量�Q�这些特征向量就可以构成�q�个投媄(ji��ng)矩阵�?ji��n)。特征向量的选择取决于协方差矩阵的特征值的大小�?/p>

举一个例子：(x��)

对于一个训�l�集�Q?00个sample�Q�特征是10�l�_(d��)��那么它可以徏立一�?00*10的矩阵，作�ؓ(f��)��h��。求�q�个��h��的协方差矩阵�Q�得��C��?0*10的协方差矩阵�Q�然后求�?gu��)��个协方差矩阵的特征值和特征向量�Q�应该有10个特征值和特征向量�Q�我们根据特征值的大小�Q�取前四个特征值所对应的特征向量，构成一�?0*4的矩阵，�q�个矩阵��是我们要求的特征矩阵，100*10的样本矩阵乘?sh��)��这�?0*4的特征矩阵，��得��C��(ji��n)一�?00*4的新的降�l�之后的��h��矩阵�Q�每个sample的维��C��降了(ji��n)�?/p>

当给定一个测试的特征集之后，比如1*10�l�的特征�Q�乘?sh��)��上面得到�?0*4的特征矩阵，便可以得��C��?*4的特征，用这个特征去分类�?/p>

所以做PCA实际上是求得�q�个投媄(ji��ng)矩阵�Q�用高维的特征乘?sh��)��这个投��q��阵，便可以将高维特征的维��C��降到指定的维数�?/p>

在opencv里面有专门的函数�Q�可以得到这个这个投��q��阵（特征矩阵�Q��?/p>

void cvCalcPCA( const CvArr* data, CvArr* avg, CvArr* eigenvalues, CvArr* eigenvectors, int flags );

polly 2012-02-02 13:24 发表评论

polly — Sat, 24 Sep 2011 14:14:00 GMT

1#include <stdio.h>
2#include <math.h>
3#include <stdlib.h>
4/*
50.500000 1.100000 3.100000
60.000000 -10.040000 -24.500000
70.000000 0.000000 -12.284024
8-2.600000
91.000000
102.000000
11��h��L��键��(h��)�l? . .
12*/
13#define n 3
14double a[3][3]={{0.5,1.1,3.1},{5,0.96,6.5},{2,4.5,0.36}};
15double b[3]={6,0.96,0.02};
16double m[3][3];
17double x[3]={0,0,0};
18int k,p;
19
20int main(){
21    for(k=0;k<n-1;k++){
22            if(a[k][k]==0)  break;
23            else{
24                     for(int i=k+1;i<n;i++){
25                             m[i][k]=a[i][k]/a[k][k];
26                             for(int j=k;j<n;j++){
27                             a[i][j]=a[i][j]-m[i][k]*a[k][j];
28                             }
29                             b[i]=b[i]-b[k]*m[i][k];
30                     }
31            }
32    }
33    for(int i=0;i<n;i++){
34            for(int j=0;j<3;j++){
35            printf("%f ",a[i][j]);}
36            printf("\n");
37    }
38    //huidai--有错
39    x[n-1]=b[n-1];//Right..
40    for(int i=n-1;i>=0;i--){//where is the error
41          double sum=0;//ok.Vectory.
42          for(int j=i+1;j<n;j++){//Be careful while using for loop!!!
43                sum+=a[i][j]*x[j];
44          }
45          x[i]=(b[i]-sum)/a[i][i];
46    }
47
48    for(int i=0;i<n;i++){
49            printf("%3.2f\n",x[i]);
50    }
51
52    system("pause");
53}
54

polly 2011-09-24 22:14 发表评论