add doc of search-pic-demo #1075
Conversation
| 以下为以图（文）搜图的流程图： | ||
|
|
||
| <div align="center"> | ||
| <img src=https://community-shared-data-1308875761.cos.ap-beijing.myqcloud.com/artwork/docs/tutorial/Vector/search-image.png width=80% heigth=80%/> |
There was a problem hiding this comment.
Please upload the image to https://github.com/matrixorigin/artwork, and then update the URL to https://github.com/matrixorigin/artwork/blob/main/docs/tutorial/Vector/search-image.png.
Ditto for img_search.png and text_search_pic.png below.
| - 下载安装 `pymysql` 工具。使用下面的代码下载安装 `pymysql` 工具： | ||
|
|
||
| ``` | ||
| pip3 install pymysql |
There was a problem hiding this comment.
Please be consistent with pip and pip3:
- either use
piphere, or - change
pipbelow topip3
| 连接 MatrixOne，建立一个名为 `pic_tab` 的表来存储图片路径信息和对应的向量信息。 | ||
|
|
||
| ```sql | ||
| create table pic_tab(pic_path varchar(200),embedding vecf64(512)); |
|
|
||
| ### 遍历图片路径 | ||
|
|
||
| 定义方法 `find_img_files` 遍历本地图片文件夹，这里我预先在本地存了苹果、香蕉、蓝莓、樱桃、杏子五种类别的水果图片，每种类别若干张，格式都为。jpg。 |
| for file in files: | ||
| if file.lower().endswith('.jpg'): | ||
| full_path = os.path.join(root, file) | ||
| img_files.append(full_path) #构建完整的文件路径 |
There was a problem hiding this comment.
nit: add one more whitespace before and after #.
Ditto for the comments for storage_img() below.
Btw, black is recommended for formatting Python code.
| return img_files | ||
| ``` | ||
|
|
||
| - 图像向量化并存入 MO |
There was a problem hiding this comment.
nit: how about replacing MO as MatrixOne (以下简称 MO )
| img_features = img_features .detach().tolist() #分离张量，转换为列表 | ||
| embeddings = img_features [0] | ||
| insert_sql = "insert into pic_tab(pic_path,embedding) values (%s, normalize_l2(%s))" | ||
| data_to_insert = (file_path, str(embeddings )) |
There was a problem hiding this comment.
nit: remove extra whitespaces, i.e., = in two assignments above and str(embeddings ).
| def create_idx(n): | ||
| cursor.execute('SET GLOBAL experimental_ivf_index = 1') | ||
| create_sql='create index idx_pic using ivfflat on pic_tab(embedding) lists=%s op_type "vector_l2_ops"' | ||
| cursor.execute(create_sql,n) |
There was a problem hiding this comment.
nit:
- remove the extra whitespace before
lists - add a whitespace btw
create_sql,andn
| if img_path: | ||
| result_path = [img_path]+[path for path_tuple in data for path in path_tuple] | ||
| else: | ||
| result_path = [path for path_tuple in data for path in path_tuple] |
There was a problem hiding this comment.
nit: the indentation is NOT consistent w/ other places.
Keep comments below consistent so that there are two whitespaces after any code, and one whitespace before the actual comment.
| show_img(img_path,1,4) | ||
| #text = ["Banana"] | ||
| #text_search_img(text,3) | ||
| #show_img(None,1,3) |
There was a problem hiding this comment.
Add the comment Uncomment below for text_search_img
|
thanks for your nice suggestions! @zuyu Make necessary changes please. @yangj1211 |
What type of PR is this?
Which issue(s) this PR fixes:
issue #1052
What this PR does / why we need it: