---
# System prepended metadata

title: Cell growth prediction model(T-08)

---

# Cell growth prediction model(T-08)
----

## ==2021/6/30 Meeting 1==
---
### Job Description
* Spec pic
![](https://i.imgur.com/5JFYHif.png =x350)

---
### schedule time
* WorkFlow chart
![](https://i.imgur.com/y2IAwKw.png =x350)

---

### Step-by-Step Function

#### ==Part1優化影像==

![](https://i.imgur.com/1qogVNm.png =x400)

---

#### ==Part2建立允收標準==
![](https://i.imgur.com/FVVrQvJ.png =x450)


---
## 開會
* CV訓練分辨 漂浮型+貼附型(Computer vision distinguish cells) 
* Traning Data
(https://github.com/basveeling/pcam)
![](https://i.imgur.com/7BeBOq2.png =x250)
* 架構:==MobilenetV2==
     * [ linear bottleneck + inverted residual--->準確度效能皆可以更加提高](https://medium.com/ai-academy-taiwan/efficient-cnn-%E4%BB%8B%E7%B4%B9-%E4%BA%8C-mobilenetv2-7809721f0bc8)
      
     ![](https://i.imgur.com/QIilmaA.png)
     
     ![](https://i.imgur.com/qAH47YY.png)


---
## 效果
* 原圖
![](https://i.imgur.com/cusRPf3.png)
* 去背後
![](https://i.imgur.com/ivpnMEM.png)
----
* [優化參考論文](https://www.ijert.org/research/breast-cancer-cell-detection-using-digital-image-processing-IJERTV1IS9046.pdf)
* [SVM K-means](https://www.nature.com/articles/s41598-020-76670-6.pdf)

去粒子
---
[udacity 深度學習課程](https://classroom.udacity.com/courses/ud187/lessons/6d543d5c-6b18-4ecf-9f0f-3fd034acd2cc/concepts/2b6512b8-a5c6-4c2e-a3b7-b63abb27df97)

## 用flask deploy model
[Deploy Machine Learning Models using Flask](https://www.youtube.com/watch?v=0nr6TPKlrN0)
[falsk 做深度學習開發](https://haosquare.com/tbrain-tomofun-audio-classification/?fbclid=IwAR3GZOtD5nBhGJpm0w0Ld4_A7I5DLzoh0LLz1oR_tDYCR8hlWz2v-JGio5c)