Large Model Prompt Segmentation

Function: Uses the Segment Anything model, inputs boxes or points prompts for image segmentation. This model relies on the segment-anything python library. If not installed, it needs to be installed from Qianyi’s pypi source using pip install segment-anything.

Input Parameters:

Name	Type	Valid Range	Default Value	Meaning
Image	ColorImage	None	None	Input image, requires a color image with RGB channels.
List of hint points	List	None	None	(np.ndarray or None): Nx2 array, each line of the array represents a hint pixel position (X, Y).
Tooltip list	List	None	None	(np.ndarray or None): Nx4x2 array, each row of the array represents the coordinates of the four points of a prompt box, and its upper left corner and lower right corner will be calculated according to this calculation.

Name

Type

Valid Range

Default Value

Meaning

Image

ColorImage

None

Input image, requires a color image with RGB channels.

List of hint points

List

None

(np.ndarray or None): Nx2 array, each line of the array represents a hint pixel position (X, Y).

Tooltip list

List

None

(np.ndarray or None): Nx4x2 array, each row of the array represents the coordinates of the four points of a prompt box, and its upper left corner and lower right corner will be calculated according to this calculation.

Output Parameters:

Name	Type	Valid Range	Default Value	Meaning
Detection results	DetectInstance	None	{}	Returns bounding boxes, categories, scores and polygon.

Name

Type

Valid Range

Default Value

Meaning

Detection results

DetectInstance

None

{}

Returns bounding boxes, categories, scores and polygon.

Parameter Settings:

Name	Type	Valid Range	Default Value	Meaning
Weight files	File	['.pth']	None	Model file. The model needs to correspond to the model type. The official model download paths are: Base model: http://10.10.10.98:9000/inference/sam/sam_vit_b_01ec64.pth, Large model: http://10.10.10.98:9000/inference/sam/sam_vit_l_0b3195.pth, Huge model: http://10.10.10.98:9000/inference/sam/sam_vit_h_4b8939.pth
Model type	String	['vit_b', 'vit_l', 'vit_h']	vit_b	Model type.
Enable GPU	Bool	None	False	Set whether to use the GPU for inference, if it is turned on, the computer needs to have a graphics card environment.
Confidence threshold	Float	[0.0, 1.0]	0.5	Detection threshold.
Output multiple results	Bool	None	False	For each prompt, whether to output multiple results, if true, each prompt outputs 3 masks, if false, only one is output.
Extra class name	String	['0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '10', '11', '12', '13', '14', '15', '16', '17', '18', '19', '20', '21', '22', '23', '24', '25', '26', '27', '28', '29']	0	You can set the category name according to the scene to perform operations such as filtering.

Name

Type

Valid Range

Default Value

Meaning

Weight files

File

['.pth']

None

Model file. The model needs to correspond to the model type. The official model download paths are: Base model: http://10.10.10.98:9000/inference/sam/sam_vit_b_01ec64.pth, Large model: http://10.10.10.98:9000/inference/sam/sam_vit_l_0b3195.pth, Huge model: http://10.10.10.98:9000/inference/sam/sam_vit_h_4b8939.pth

Model type

String

['vit_b', 'vit_l', 'vit_h']

vit_b

Model type.

Enable GPU

Bool

None

False

Set whether to use the GPU for inference, if it is turned on, the computer needs to have a graphics card environment.

Confidence threshold

Float

[0.0, 1.0]

0.5

Detection threshold.

Output multiple results

Bool

None

False

For each prompt, whether to output multiple results, if true, each prompt outputs 3 masks, if false, only one is output.

Extra class name

String

['0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '10', '11', '12', '13', '14', '15', '16', '17', '18', '19', '20', '21', '22', '23', '24', '25', '26', '27', '28', '29']

You can set the category name according to the scene to perform operations such as filtering.