Hi,<div><br></div><div>This week <span class="Apple-style-span" style="font-family: Helvetica, Arial, sans-serif; font-size: 13px; line-height: 17px; ">I begin to implement the image registration function for astronomical images.</span></div>
<div><span class="Apple-style-span" style="font-family: Helvetica, Arial, sans-serif; font-size: 13px; line-height: 17px; "><br></span></div><div><span class="Apple-style-span" style="font-family: Helvetica, Arial, sans-serif; font-size: 13px; line-height: 17px; ">Images taken at different times may have relative translation and rotations. </span><span class="Apple-style-span" style="font-family: Helvetica, Arial, sans-serif; font-size: 13px; line-height: 17px; ">For astronomical images, stars are ideal reference points for images, because they are both plentiful and fixed in space. Stars are easily to be automatically located in the image. Also, even with translations and rotations, the relative distance between two stars shall remain the same.</span></div>
<span class="Apple-style-span" style="font-family: Helvetica, Arial, sans-serif; font-size: 13px; line-height: 17px; "><div><span class="Apple-style-span" style="font-family: Helvetica, Arial, sans-serif; font-size: 13px; line-height: 17px; "><br>
</span></div>Based on the above reason, my image registration algorithm is as follows:</span><span class="Apple-style-span" style="font-family: Helvetica, Arial, sans-serif; font-size: 13px; line-height: 17px; "><p style="font-size: 10pt; line-height: 13pt; color: rgb(0, 0, 0); font-weight: normal; margin-top: 10px; margin-right: 0px; margin-bottom: 10px; margin-left: 0px; padding-top: 0px; padding-right: 0px; padding-bottom: 0px; padding-left: 0px; ">
First, we search and locate the appropriate stars in the master image. Then for each located star, we calculate and record the distances between current star an all the other located stars. Let D<sub>ij</sub> denotes the distance between star i and j in the master image. Then we do the same operation on the slave image. Let S<sub>ij</sub> denotes the distance between star i and j in the slave image.</p>
<p style="font-size: 10pt; line-height: 13pt; color: rgb(0, 0, 0); font-weight: normal; margin-top: 10px; margin-right: 0px; margin-bottom: 10px; margin-left: 0px; padding-top: 0px; padding-right: 0px; padding-bottom: 0px; padding-left: 0px; ">
Then we shall perform a matching between the located stars in the master image and stars in the slave image based on the relative distances between the star and other stars.</p></span><div><span class="Apple-style-span" style="font-family: Helvetica, Arial, sans-serif; font-size: 13px; line-height: 17px; "><br>
</span></div><div><span class="Apple-style-span" style="font-family: Helvetica, Arial, sans-serif; font-size: 13px; line-height: 17px; ">Next week I will finish the implementation of image registration. I shall calculate the rotation and translation parameters based on the matching stars and align the two images together.</span></div>
<div><span class="Apple-style-span" style="font-family: Helvetica, Arial, sans-serif; font-size: 13px; line-height: 17px; "><br></span></div><div><span class="Apple-style-span" style="font-family: Helvetica, Arial, sans-serif; font-size: 13px; line-height: 17px; ">Detailed report can be found here: </span><a href="http://opticks.org/confluence/display/~yiwei/Status+Report+for+Week+12%2C+GSoC+2011">http://opticks.org/confluence/display/~yiwei/Status+Report+for+Week+12%2C+GSoC+2011</a></div>