Response Parameters

IsSuccess

Mandatory

Bool

Specifies whether the inference is successful.

ErrorMsg

Optional

String

Inference error message.

ErrorCode

Optional

String

Inference error code.

Result

Mandatory

String

OK/NG.

InferenceTime

Mandatory

Object of Table 2 Inference Time

Time consumed by request, inference, and postprocessing, in seconds.

Objects

Optional

Array of Table 3 Object

Component error/missing/reverse scenarios: returned box and category information.

SubObjects

Optional

Array of Table 3 Object

Returned level-2 detection object set. The coordinates of the box are those of the original image.

Patterns

Optional

Table 4

Component error/missing/reverse scenarios: template information during image registration.

Texts

Optional

Table 7

Array of Object

Industrial OCR scenario: returned inference information.

BlobMask

Optional

String

Gel detection scenario: encoded Blob Mask image. The content in the image is the segmented mask image, indicating the Blob ID of each pixel.

LabelMap

Optional

Object

Glue detection scenario: mapping between segmenting digital labels and label names.

BlobContours

Optional

String

Glue detection scenario: Blob contour image after Base64 encoding.

Blobs

Optional

Table 8

Array of Objects

Glue detection scenario: detailed information about the Blob ID area.

MissedBlobs

Optional

Table 8

Array of Objects

Gel detection scenario: missing Blob ID information.

PodId

Optional

String

Gel detection scenario: Pod ID.

PodsNum

Optional

Number

Gel detection scenario: Pods Number.

RegisterMatrix

Optional

Array of Number

Returned rotation offset matrix in WithReg mode.

In the error/missing/reverse detection scenarios:

• If the RegisterMatrix is passed, use this rotation offset matrix to complete registration.
• If the RegisterMatrix is not passed, the system calculates an optimal rotation-translation matrix for registration based on the passed patterns and bounding boxes. Note that the number of detection objects must be greater than or equal to 2.

When the number of paired template boxes and bounding boxes is less than two, the built-in 2D registration function of the system is not supported. In this case, the vision software needs to use other registration algorithms and pass the calculated correct RegisterMatrix field in the request parameters.

Prediction

Optional

objects

Semiconductor quality inspection scenario: prediction results.

All

Mandatory

Float

Time consumed by sending a request for inference and returning the inference result, in seconds.

PostAI

Mandatory

Float

Time consumed by postprocessing, in seconds.

Stream

Mandatory

Float

Time consumed by model inference, in seconds.

StartTime

Mandatory

Float

Inference start time, in the format of YYYYMMDDHHMMSS.decimals.

EndTime

Mandatory

Float

Inference end time, in the format of YYYYMMDDHHMMSS.decimals.

Box

Mandatory

Table 4

Bounding box.

Classification

Mandatory

Objects of Table 5

Secondary classification result.

Label

Mandatory

String

Category of the bounding box area.

Score

Mandatory

Float

Confidence of the bounding box.

result

Mandatory

String

Postprocessing result. The value can be OK or NG.

Segmentation

Optional

Object

Segmentation result. This parameter is left empty by default in the component error/missing/reverse scenarios.

OCR

Optional

Object

OCR recognition result. This parameter is left empty by default in the component error/missing/reverse scenarios.

X

Mandatory

Integer

Value of the X coordinate of a rectangle.

Y

Mandatory

Integer

Value of the Y coordinate of a rectangle.

Width

Mandatory

Integer

Width of a rectangle.

Height

Mandatory

Integer

Height of a rectangle.

Angle

Mandatory

Integer

Angle of a rectangle.

Label

Mandatory

String

Classification type.

Score

Mandatory

Float

Classification confidence.

Box

Mandatory

Objects of Table 4

Width of a rectangle.

Area

Mandatory

Integer

Area of a split region.

ID

Mandatory

Integer

Pixel value of a split region.

Label

Mandatory

String

Category of a split region.

result

Mandatory

String

Result of the registration Blob area. The value can be OK or NG.

OCRText

Mandatory

String

Recognized text information.

Quad

Mandatory

Objects of Table 8 Quad

Coordinates of the text box.

Score

Mandatory

Float

Recognition confidence.

BottomLeft

Mandatory

Objects of Table 9

Values of X and Y coordinates in the lower left corner.

BottomRight

Mandatory

Objects of Table 9

Values of X and Y coordinates in the lower right corner.

TopLeft

Mandatory

Objects of Table 9

Values of X and Y coordinates in the upper left corner.

TopRight

Mandatory

Objects of Table 9

Values of X and Y coordinates in the upper right corner.

X

Mandatory

Float

Value of X coordinate.

Y

Mandatory

Float

Value of Y coordinate.

bincode

Mandatory

String

The value 00 indicates success is returned, and the value 01 indicates reject is returned.

center

Mandatory

Array

Coordinates of the center point of the inference image in the corresponding reference image.

defect_point

Mandatory

Array

Location of the exception point. The first two values indicate the coordinates of the upper left corner of the box, and the last two values indicate the coordinates of the lower right corner of the box.

model_version

Mandatory

String

Inference model version information.

stat

Mandatory

Integer

Preprocessing for an image to be inferred. The value 1 indicates success, and the value 0 indicates failure. If the operation is successful, model inference can be performed. If the operation fails, an error code is returned.

blur_weight

Mandatory

Float

Blurriness of an image to be inferred during image preprocessing. If the blurriness of the image does not meet the threshold requirement, an error code is returned.

show_prob

Mandatory

Float

Predicted value of the model. If the value is greater than the specified threshold, reject is returned.