I modeled after completing the camera function, but the final model is only half of the original object, such as a pair of shoes. I obviously have corresponding photos at 360 degrees, but there are only 180-degree objects in Reality Capture. At the same time, the photo on the other side cannot be detected by the camera modell
The most common problem with the images not connecting is the lack of common features. This could be because of the changing shooting conditions, or maybe insufficient overlap.
If you believe that the overlap is sufficient, I would recommend using control points:
Control Point Workflow
This video has the old UI, but the control point placing principle is the same.
Create 4-5 control points and assign them to up to 10 images (from both sides) and then run alignment.
Nah, I already set the control points,But like the picture it didnt hat right side camera sensor.So i set the control points with the right side pictures,but still useless
Did you happen to use a turn table setup with a non uniform white or black background. I have seen errors like this when people attempt to take pictures from a static position with a turling object, while keeping the background as is, meaning not setting it up, so that everything other than the turn table and object to scan is near pure white or near pure black. If you use a turn table, you must make the background as featureless as possible in order to trick RC into thinking that it was the camera that moved around the object.
I didnt use a trun table to take my photos.I did everything maunell and in order to get more angle for different points in object,I tried 360 degrees to build my pictures.There are white walls on both sides（which higher different features to my black shoes） and express boxes(like normal yellow color ) below my shoes
Hi giraffe_zhou,
how many images do you have in that project? Are the images taken in more rows around the shoe (as a globe)? What is your overlap between images? What were your alignment settings? Do you receive more components after alignment? As you are using control points, on how many images are those placed (one point on how many images)? If you have more components, are control points placed on both components?
On your print screen there is no visible placed control point on your model. Have you turned the showing option off? If not, if is possible that CPs are placed separately on just one image each.
how many images do you have in that project? Are the images taken in more rows around the shoe (as a globe)? What is your overlap between images? What were your alignment settings? Do you receive more components after alignment? As you are using control points, on how many images are those placed (one point on how many images)? If you have more components, are control points placed on both components?
On your print screen there is no visible placed control point on your model. Have you turned the showing option off? If not, if is possible that CPs are placed separately on just one image each.
Hi, we will need to see your data and also project. Can you contact our support at support@capturingreality.com and send us your data and your project (project file and project folder).