This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Vision library function in Candy edge detection (VLIB_xyGradientsAndMagnitude)

Hello, everyone.

         I want to use the VLIB in my DM642. First,  I just use the function "VLIB_nonMaximumSuppressionCanny" for a test, and it work well.

         But when I use the function  "VLIB_xyGradientsAndMagnitude( imageData,  gradX + imageWidth + 1,  gradY + imageWidth + 1,  gradM + imageWidth + 1,  imageWidth,  imageHeight);",   I find the out put result "gradX, gradY, gradM" are wrong. Some of them are all zero.

         I find the same problem in this website:http://e2e.ti.com/support/dsp/davinci_digital_media_processors/f/99/t/5803.aspx , I saw the reply"This issue is fixed in the new release. It should be on the website already or in a couple of days."  But I got this lib  "VLIB2.0"   this year,  I don't know whether it is new enough. I use a demo and a picture size 128x128 for a test in CCS4 simulator. And it work well. Then I chose the others size as follow: 240x320  512x512  480x640, and they all work well too. But when I use 576x720, its result become wrong.   

         The video frame  captured and display in my program in DM642 is size 576x720.  How can I solve my problem? Is there anything or detail I should be care but I miss?

        This is my first time discuss problem in this forums, and  I have been stuck by this problem for a few days. 

        I really hope anyone can give me some advice or help me, I will really appreciate it, and thank you very much!!