Lyndon Bermoy Active Member Licensed User Longtime User Feb 25, 2016 #1 Hi guys! I just want to ask if it is possible to have image recognition processing? For example, I will capture a tomato in my camera and then after capturing, it will inform the user that it is a tomato and showing information. But the main question is, is there really a way to match the capturing image to the real world image such as vegetables identification? Please I really need an answer.. thanks
Hi guys! I just want to ask if it is possible to have image recognition processing? For example, I will capture a tomato in my camera and then after capturing, it will inform the user that it is a tomato and showing information. But the main question is, is there really a way to match the capturing image to the real world image such as vegetables identification? Please I really need an answer.. thanks
JordiCP Expert Licensed User Longtime User Feb 25, 2016 #2 My one-and-a-half cent: If you want to detect "whatever", I would advise to capture the image and send it to an online server which sends you back the results. Seems that Google has its own service for this https://cloud.google.com/vision/ , don't know how their API works and also there are other solutions, some of them http-based (simpler to manage), but no idea of their performance nor if they are free/paid http://www.cloudsightapi.com/ https://www.quora.com/Which-company-has-the-best-image-recognition-APIs-in-the-market-place-today Upvote 0
My one-and-a-half cent: If you want to detect "whatever", I would advise to capture the image and send it to an online server which sends you back the results. Seems that Google has its own service for this https://cloud.google.com/vision/ , don't know how their API works and also there are other solutions, some of them http-based (simpler to manage), but no idea of their performance nor if they are free/paid http://www.cloudsightapi.com/ https://www.quora.com/Which-company-has-the-best-image-recognition-APIs-in-the-market-place-today
lemonisdead Well-Known Member Licensed User Longtime User Feb 25, 2016 #3 Hello, Lyndon Bermoy said: I just want to ask if it is possible to have image recognition processing? Click to expand... Yes it is possible but as far as I know we don't have a native library and should use an external API relying on paid services (I have done something in the past as a test drive with MoodStocks but they are a lot of providers). Lyndon Bermoy said: But the main question is, is there really a way to match the capturing image to the real world image such as vegetables identification? Click to expand... My tests worked not so bad but were correct at 40% of time only (not comparable with Googles app recognition ones) Upvote 0
Hello, Lyndon Bermoy said: I just want to ask if it is possible to have image recognition processing? Click to expand... Yes it is possible but as far as I know we don't have a native library and should use an external API relying on paid services (I have done something in the past as a test drive with MoodStocks but they are a lot of providers). Lyndon Bermoy said: But the main question is, is there really a way to match the capturing image to the real world image such as vegetables identification? Click to expand... My tests worked not so bad but were correct at 40% of time only (not comparable with Googles app recognition ones)