Recent study schedule

Mainly divided into three points

1. First, I carefully read several innovative papers on Cascade AdaBoost. I have read three articles and the remaining six foreign documents.

2. Take a look at C ++ and Ajax. It seems that you can only learn more about projects, or you will forget everything you have read.

3. In terms of identification of Citrus ulcer disease, another month's foreign child failed. I don't know where to start? Improvements needed

A features: adds color features, simple texture features, retests Gabor features, deletes Fourier features, and edge detection operator features.

B window sample: add other disease samples and other background samples (such as tree trunk) to the negative sample)

Increase the number of samples

C sample of lesions: Same as above

D. How to increase the speed?