1
0
mirror of https://github.com/opencv/opencv_contrib.git synced 2025-10-24 20:01:12 +08:00
Commit Graph

2900 Commits

Author SHA1 Message Date
berak
00ea0f49df saliency: improve BING handling 2017-10-23 17:56:35 +02:00
cDc
4095ef10db Fix name extraction from path in multicalib 2017-10-23 16:04:01 +03:00
Vladislav Sovrasov
0c5a8e1fec tracking: speedup NCC kernel for TLD and Median Flow 2017-10-23 15:15:53 +03:00
Vladislav Sovrasov
941865c9a1 tracking: reduce useless computations in TLD 2017-10-23 15:15:53 +03:00
Vladislav Sovrasov
62939e2937 tracking: slightly rewrite main loop in TLD 2017-10-23 15:15:53 +03:00
Vladislav Sovrasov
a0896c7bd1 tracking: update tresholds in regression tests of the TLD 2017-10-23 15:15:53 +03:00
Vladislav Sovrasov
895a9143a6 tracking: fix unitialized memory access in TLD 2017-10-23 15:15:53 +03:00
Vladislav Sovrasov
a6215264ed tracking: eliminate code duplication 2017-10-23 15:15:53 +03:00
Vladislav Sovrasov
b8588f845f tracking: disable OCL in TLD (kernels are slow and broken) 2017-10-23 15:15:53 +03:00
Vladislav Sovrasov
ccfd55dc7a tracking: enable tests for TLD 2017-10-23 15:15:53 +03:00
Vladislav Sovrasov
e83871228c TLD: fix NCC calculation 2017-10-23 15:15:53 +03:00
Vladislav Sovrasov
255de8777b tracking: fix wrong conditions of OCL run 2017-10-23 15:15:53 +03:00
Vladislav Sovrasov
926b7af8c9 tracking: hide unused debug method in MedianFlow 2017-10-23 15:15:53 +03:00
Vitaly Tuzov
a9b8fb8300 GaussianBlur of the source image for LATCH descriptor made optional 2017-10-20 21:48:16 +03:00
Maksim Shabunin
ca8adc9a49 Removed obsolete contrib_world module 2017-10-19 15:55:24 +03:00
Maksim Shabunin
0bd061f07b Merge pull request #1413 from alalek:fix_build 3.3.1 2017-10-18 14:51:04 +00:00
Pavel Rojtberg
4de2b67e71 aruco: use std::rotate to avoid copying all corners 2017-10-18 15:38:49 +02:00
Alexander Alekhin
27ced2cb2c tracing: disable unused code 2017-10-18 13:06:50 +03:00
Alexander Alekhin
fe7842e8cb drop unused code 2017-10-18 13:01:13 +03:00
Alexander Alekhin
a22b9815b1 fix code style 2017-10-18 13:00:28 +03:00
Alexander Alekhin
5ba8e40a53 surface_matching: fix infinite loop, whitespace issues 2017-10-18 12:55:00 +03:00
Alexander Alekhin
11c90ea23c cvv: fix build 2017-10-18 12:51:24 +03:00
Alexander Alekhin
bdf51c3a2b contrib: fix protobuf usage 2017-10-18 12:01:05 +03:00
Alexander Alekhin
e9cd99d9f7 Merge pull request #1406 from alalek:matlab_module 2017-10-13 15:23:14 +00:00
Vitaly Tuzov
220edb81bc Updated 32F disparity matrix handling 2017-10-13 15:09:00 +03:00
Alexander Alekhin
26fd198ef2 matlab: disable module by default
- broken bindings generator
- broken conversion functions (uses non-supported cv::Mat types)
- other problems like missing convertion functions for Moments and other structs
2017-10-13 12:22:05 +03:00
Vladislav Sovrasov
27961cd8cc text: fix wrong channel swap in TestDetectorCNN 2017-10-12 18:29:30 +03:00
Vladislav Sovrasov
7031316cb7 text: add text recognition sample 2017-10-11 14:48:35 +03:00
Vladislav Sovrasov
9195d2e614 text: small adjustments in samples and image preprocessing 2017-10-11 14:47:52 +03:00
Vladislav Sovrasov
fb0338fb92 Merge branch 'master' into text_detector_dnn 2017-10-11 12:29:59 +03:00
Hamdi Sahloul
2cfc353172 Re-write surface matching using vectors and matrices 2017-10-10 20:27:09 +03:00
Vladislav Sovrasov
3253fe9f7e text: impovements in samples and module interface 2017-10-10 16:08:35 +03:00
Alexander Alekhin
374e321877 build warning (#1400) 2017-10-10 15:55:46 +03:00
Vladislav Sovrasov
1306621f3d text: add prototxt for text detection model 2017-10-10 15:29:20 +03:00
Vladislav Sovrasov
951e18272d text: cleanup dnn text detection part 2017-10-10 14:16:13 +03:00
Vladislav Sovrasov
41995b76e8 KCF speedup (#1374)
* kcf use float data type rather than double.

In our practice, float is good enough and could get better performance.
With this patch, one of my benchmark could get about 20% performance gain.

Signed-off-by: Zhigang Gong <zhigang.gong@intel.com>

* Offload transpose matrix multiplication to ocl.

The matrix multiplication in updateProjectMatrix is one of the
hotspot. And because of the matrix shape is special, say the
m is very short but the n is very large. The GEMM implementation
in neither the clBLAS nor the in trunk implementation are very
inefficient, I implement an standalone transpose matrix mulplication
kernel here. It can get about 10% performance gain on Intel
desktop platform or 20% performance gain on a braswell platform.
And in the mean time, the CPU utilization will be lower.

Signed-off-by: Zhigang Gong <zhigang.gong@intel.com>

* Add verification code for kcf ocl transpose mm kernel.

Signed-off-by: Zhigang Gong <zhigang.gong@linux.intel.com>

* tracking: show FPS in traker sample

* tracking: fix MSVC warnings in KCF

* tracking: move OCL kernel initialization to constructor in KCF
2017-10-10 13:54:22 +03:00
Gregor Kovalčík
0058eca130 Fixed PCT Signature typos. (#1346) 2017-10-10 13:51:36 +03:00
atinfinity
a1b3c84c5a updated aruco tutorial (#1263) 2017-10-10 13:50:19 +03:00
Vadim Pisarevsky
68736a2ce5 Merge pull request #1384 from mshabunin:pr723 2017-10-10 09:54:46 +00:00
Vadim Pisarevsky
8fbb0ec8eb Merge pull request #1387 from terfendail:fast_extension 2017-10-09 13:24:38 +00:00
Vadim Pisarevsky
47771c1806 Merge pull request #1390 from sovrasov:saliency_update 2017-10-09 13:23:41 +00:00
Vadim Pisarevsky
9b275d58da Merge pull request #1393 from terfendail:surface_matching_ply 2017-10-09 13:19:19 +00:00
Vladislav Sovrasov
c33629e053 Merge branch 'master' into text_detector_dnn 2017-10-09 14:38:20 +03:00
Vitaly Tuzov
a922dd5976 updated loadPLYSimple automatic detection of normals, color, and alpha properties 2017-10-04 13:40:25 +03:00
Andrew Hundt
4e6e8a4e2a loadPLYSimple basic automatic detection of normals, color, and alpha properties 2017-10-04 08:54:34 +03:00
Andrew Hundt
47237bf115 surface_matching loadPLYSimple() reports only ascii ply is supported (+1 squashed commit)
Squashed commits:
[802ce32] surface_matching loadPLYSimple() reports only ascii ply is supported (+1 squashed commit)
Squashed commits:
[e240207] surface_matching loadPLYSimple() reports only ascii ply is supported
2017-10-03 17:14:13 +03:00
Vladislav Sovrasov
e1cd60c96e saliency: remove broken perf tests 2017-10-03 16:57:55 +03:00
Vladislav Sovrasov
cf68a01138 saliency: cleanup module and samples 2017-10-02 17:31:36 +03:00
jaco
3f4e7dfc10 ceil explicit cast added 2017-10-02 16:30:46 +03:00
jaco
83f80a34ca windows warning wip 2017-10-02 16:30:46 +03:00