Hello, it is well known that the images collected by cameras with different pixels have different sizes. If the size of the image I used for training and the size of the image used for recognition are inconsistent, for example, my training set is collected with a 12-megapixel camera , But the images I use for recognition are all captured with a 3 megapixel camera, so is there any problem? Will it be impossible to achieve a good recognition effect because the size of the anchor is inconsistent with the entire image? Looking forward to your answer, thank you.
Hello, it is well known that the images collected by cameras with different pixels have different sizes. If the size of the image I used for training and the size of the image used for recognition are inconsistent, for example, my training set is collected with a 12-megapixel camera , But the images I use for recognition are all captured with a 3 megapixel camera, so is there any problem? Will it be impossible to achieve a good recognition effect because the size of the anchor is inconsistent with the entire image? Looking forward to your answer, thank you.