(see literature research of street landscape) ran across network SegNet image segmentation algorithm, and feel it recognition effect is very ideal, follow the documents found at the university of Cambridge ask SegNet platform (and I don't know whether I call platform, is a web page, you can directly import the images and the results of that kind of), and found that this platform can't identify I upload pictures, with my poor knowledge of computers, estimation is my pictures (natural forests in the view of the people) and machine learning training gallery (is) the street landscape is not consistent?
And then fell into deep confusion,,, I am not a computer professional, artificial intelligence, machine learning, what you see is only slightly background knowledge is not enough to support my further looking for other ways, since the turn to the great god omnipotent computer:
1 ever learn the image type, can direct application platform or program ?
2 if there is no such a platform, is there any other method of image recognition?
Impressed by the rapid development of science and technology, as a will only use the final results of the small white, also hope that as much as possible in their own professional study using high-tech to improve the rigor of research, sincere hope, every answer, thank!
Ps. I really don't know which knowledge, in the field of image segmentation is the ball plate if bosses light spray
