fastdeploy::vision Namespace Reference

All C++ FastDeploy Vision Models APIs are defined inside this namespace. More...

Namespaces
	classification
	All object classification model APIs are defined inside this namespace.
	detection
	All object detection model APIs are defined inside this namespace.
	facedet
	All object face detection model APIs are defined inside this namespace.
	keypointdetection
	All keypoint detection model APIs are defined inside this namespace.
	matting
	All image/video matting model APIs are defined inside this namespace.
	ocr
	All OCR series model APIs are defined inside this namespace.
	segmentation
	All segmentation model APIs are defined inside this namespace.

Classes
class	BGR2GRAY
	Processor for tansform images from BGR to GRAY. More...
class	BGR2RGB
	Processor for tansform images from BGR to RGB. More...
class	Cast
	Processor for cast images with given type deafault is float. More...
class	CenterCrop
	Processor for crop images in center with given type deafault is float. More...
struct	ClassifyResult
	Classify result structure for all the image classify models. More...
class	Convert
	Processor for convert images with given paramters. More...
class	ConvertAndPermute
	Processor for convert images with given paramters and permute images from HWC to CHW. More...
class	Crop
	Processor for crop images with given paramters. More...
struct	DetectionResult
	Detection result structure for all the object detection models and instance segmentation models. More...
struct	FaceAlignmentResult
	Face Alignment result structure for all the face alignment models. More...
struct	FaceDetectionResult
	Face detection result structure for all the face detection models. More...
struct	FaceRecognitionResult
	Face recognition result structure for all the Face recognition models. More...
struct	FDMatBatch
	FDMatBatch contains batch data for preprocess. More...
struct	HeadPoseResult
	HeadPose result structure for all the headpose models. More...
class	HWC2CHW
	Processor for transform images from HWC to CHW. More...
struct	KeyPointDetectionResult
	KeyPoint Detection result structure for all the keypoint detection models. More...
class	LimitByStride
	Processor for LimitByStride images with given paramters. More...
class	LimitShort
	Processor for Limit images by short edge with given paramters. More...
struct	Mask
struct	Mat
	FDMat is a structure for replace cv::Mat. More...
struct	MattingResult
	Matting result structure for all the Matting models. More...
struct	MOTResult
	MOT(Multi-Object Tracking) result structure for all the MOT models. More...
class	Normalize
	Processor for Normalize images with given paramters. More...
class	NormalizeAndPermute
	Processor for Normalize and Permute images from HWC to CHW. More...
class	Pad
	Processor for padding images. More...
class	PadToSize
	Processor for padding images to given size. More...
class	Processor
	Processor base class for processors in fastdeploy/vision/common/processors. More...
class	ProcessorManager
	ProcessorManager for Preprocess. More...
class	Resize
	Processor for Resize images. More...
class	ResizeByShort
	Processor for resize images by short edge. More...
class	RGB2BGR
	Processor for tansform images from RGB to BGR. More...
class	RGB2GRAY
	Processor for tansform images from RGB to GRAY. More...
struct	SegmentationResult
	Segmentation result structure for all the segmentation models. More...
class	StridePad
	Processor for padding images with stride. More...

Functions
void	EnableFlyCV ()
	Enable using FlyCV to process image while deploy vision models. Currently, FlyCV in only available on ARM(Linux aarch64/Android), so will fallback to using OpenCV in other platform.
void	DisableFlyCV ()
	Disable using FlyCV to process image while deploy vision models.
void	SetProcLibCpuNumThreads (int threads)
	Set the cpu num threads of ProcLib.
cv::Mat	VisClassification (const cv::Mat &im, const ClassifyResult &result, int top_k=5, float score_threshold=0.0f, float font_size=0.5f)
	Show the visualized results for classification models. More...
cv::Mat	VisClassification (const cv::Mat &im, const ClassifyResult &result, const std::vector< std::string > &labels, int top_k=5, float score_threshold=0.0f, float font_size=0.5f)
	Show the visualized results with custom labels for classification models. More...
cv::Mat	VisDetection (const cv::Mat &im, const DetectionResult &result, float score_threshold=0.0, int line_size=1, float font_size=0.5f)
	Show the visualized results for detection models. More...
cv::Mat	VisDetection (const cv::Mat &im, const DetectionResult &result, const std::vector< std::string > &labels, float score_threshold=0.0, int line_size=1, float font_size=0.5f)
	Show the visualized results with custom labels for detection models. More...
cv::Mat	VisFaceAlignment (const cv::Mat &im, const FaceAlignmentResult &result, int line_size=1)
	Show the visualized results for face alignment models. More...
cv::Mat	VisFaceDetection (const cv::Mat &im, const FaceDetectionResult &result, int line_size=1, float font_size=0.5f)
	Show the visualized results for face detection models. More...
cv::Mat	VisKeypointDetection (const cv::Mat &im, const KeyPointDetectionResult &results, float conf_threshold=0.5f)
	Show the visualized results for key point detection models. More...
cv::Mat	VisMatting (const cv::Mat &im, const MattingResult &result, bool transparent_background=false, float transparent_threshold=0.999, bool remove_small_connected_area=false)
	Show the visualized results for matting models. More...
cv::Mat	VisOcr (const cv::Mat &im, const OCRResult &ocr_result, const float score_threshold=0)
	Show the visualized results for Ocr models. More...
cv::Mat	VisSegmentation (const cv::Mat &im, const SegmentationResult &result, float weight=0.5)
	Show the visualized results for segmentation models. More...
cv::Mat	SwapBackground (const cv::Mat &im, const cv::Mat &background, const MattingResult &result, bool remove_small_connected_area=false)
	Swap the image background with MattingResult. More...
cv::Mat	SwapBackground (const cv::Mat &im, const cv::Mat &background, const SegmentationResult &result, int background_label)
	Swap the image background with SegmentationResult. More...

Detailed Description

All C++ FastDeploy Vision Models APIs are defined inside this namespace.

Function Documentation

SwapBackground() [1/2]

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::SwapBackground	(	const cv::Mat &	im,
		const cv::Mat &	background,
		const MattingResult &	result,
		bool	remove_small_connected_area = false
	)

Swap the image background with MattingResult.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	background	the background image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	result	the MattingResult produced by model
[in]	remove_small_connected_area	if remove_small_connected_area==true, the visualized result will not include the small connected areas

Returns

cv::Mat type stores the visualized results

SwapBackground() [2/2]

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::SwapBackground	(	const cv::Mat &	im,
		const cv::Mat &	background,
		const SegmentationResult &	result,
		int	background_label
	)

Swap the image background with SegmentationResult.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	background	the background image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	result	the SegmentationResult produced by model
[in]	background_label	the background label number in SegmentationResult

Returns

cv::Mat type stores the visualized results

VisClassification() [1/2]

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::VisClassification	(	const cv::Mat &	im,
		const ClassifyResult &	result,
		int	top_k = 5,
		float	score_threshold = 0.0f,
		float	font_size = 0.5f
	)

Show the visualized results for classification models.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	result	the result produced by model
[in]	top_k	the length of return values, e.g., if topk==2, the result will include the 2 most possible class label for input image.
[in]	score_threshold	threshold for top_k scores, the class will not be shown if the score is less than score_threshold
[in]	font_size	font size

Returns

cv::Mat type stores the visualized results

VisClassification() [2/2]

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::VisClassification	(	const cv::Mat &	im,
		const ClassifyResult &	result,
		const std::vector< std::string > &	labels,
		int	top_k = 5,
		float	score_threshold = 0.0f,
		float	font_size = 0.5f
	)

Show the visualized results with custom labels for classification models.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	result	the result produced by model
[in]	labels	custom labels for user, the visualized result will show the corresponding custom labels
[in]	top_k	the length of return values, e.g., if topk==2, the result will include the 2 most possible class label for input image.
[in]	score_threshold	threshold for top_k scores, the class will not be shown if the score is less than score_threshold
[in]	font_size	font size

Returns

cv::Mat type stores the visualized results

VisDetection() [1/2]

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::VisDetection	(	const cv::Mat &	im,
		const DetectionResult &	result,
		float	score_threshold = 0.0,
		int	line_size = 1,
		float	font_size = 0.5f
	)

Show the visualized results for detection models.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	result	the result produced by model
[in]	score_threshold	threshold for result scores, the bounding box will not be shown if the score is less than score_threshold
[in]	line_size	line size for bounding boxes
[in]	font_size	font size for text

Returns

cv::Mat type stores the visualized results

VisDetection() [2/2]

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::VisDetection	(	const cv::Mat &	im,
		const DetectionResult &	result,
		const std::vector< std::string > &	labels,
		float	score_threshold = 0.0,
		int	line_size = 1,
		float	font_size = 0.5f
	)

Show the visualized results with custom labels for detection models.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	result	the result produced by model
[in]	labels	the visualized result will show the bounding box contain class label
[in]	score_threshold	threshold for result scores, the bounding box will not be shown if the score is less than score_threshold
[in]	line_size	line size for bounding boxes
[in]	font_size	font size for text

Returns

cv::Mat type stores the visualized results

VisFaceAlignment()

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::VisFaceAlignment	(	const cv::Mat &	im,
		const FaceAlignmentResult &	result,
		int	line_size = 1
	)

Show the visualized results for face alignment models.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	result	the result produced by model
[in]	line_size	line size for circle point

Returns

cv::Mat type stores the visualized results

VisFaceDetection()

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::VisFaceDetection	(	const cv::Mat &	im,
		const FaceDetectionResult &	result,
		int	line_size = 1,
		float	font_size = 0.5f
	)

Show the visualized results for face detection models.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	result	the result produced by model
[in]	line_size	line size for bounding boxes
[in]	font_size	font size for text

Returns

cv::Mat type stores the visualized results

VisKeypointDetection()

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::VisKeypointDetection	(	const cv::Mat &	im,
		const KeyPointDetectionResult &	results,
		float	conf_threshold = 0.5f
	)

Show the visualized results for key point detection models.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	results	the result produced by model
[in]	conf_threshold	threshold for result scores, the result will not be shown if the score is less than conf_threshold

Returns

cv::Mat type stores the visualized results

VisMatting()

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::VisMatting	(	const cv::Mat &	im,
		const MattingResult &	result,
		bool	transparent_background = false,
		float	transparent_threshold = 0.999,
		bool	remove_small_connected_area = false
	)

Show the visualized results for matting models.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	result	the result produced by model
[in]	transparent_background	if transparent_background==true, the background will with transparent color
[in]	transparent_threshold	since the alpha value in MattringResult is a float between [0, 1], transparent_threshold is used to filter background pixel
[in]	remove_small_connected_area	if remove_small_connected_area==true, the visualized result will not include the small connected areas

Returns

cv::Mat type stores the visualized results

VisOcr()

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::VisOcr	(	const cv::Mat &	im,
		const OCRResult &	ocr_result,
		const float	score_threshold = 0
	)

Show the visualized results for Ocr models.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	result	the result produced by model

Returns

cv::Mat type stores the visualized results

VisSegmentation()

FASTDEPLOY_DECL cv::Mat fastdeploy::vision::VisSegmentation	(	const cv::Mat &	im,
		const SegmentationResult &	result,
		float	weight = 0.5
	)

Show the visualized results for segmentation models.

Parameters

[in]	im	the input image data, comes from cv::imread(), is a 3-D array with layout HWC, BGR format
[in]	result	the result produced by model
[in]	weight	transparent weight of visualized result image

Returns

cv::Mat type stores the visualized results