Code style refactoring

Author: Nuzhny007

src/Detector/OCVDNNDetector.cpp

@@ -227,7 +227,7 @@ void OCVDNNDetector::Detect(const cv::UMat& colorFrame)
         [](const CRegion& reg) { return reg.m_brect; },
         [](const CRegion& reg) { return reg.m_confidence; },
         [](const CRegion& reg) { return reg.m_type; },
-        0, 0.f);
+        0, static_cast<track_t>(0));
 }
 
 ///

src/Detector/YoloDarknetDetector.cpp

@@ -162,11 +162,11 @@ void YoloDarknetDetector::Detect(const cv::UMat& colorFrame)
 
 		if (crops.size() > 1 || m_batchSize > 1)
 		{
-			nms3<CRegion>(tmpRegions, m_regions, 0.4f,
+			nms3<CRegion>(tmpRegions, m_regions, static_cast<track_t>(0.4),
 				[](const CRegion& reg) { return reg.m_brect; },
 				[](const CRegion& reg) { return reg.m_confidence; },
 				[](const CRegion& reg) { return reg.m_type; },
-				0, 0.f);
+				0, static_cast<track_t>(0));
 			//std::cout << "nms for " << tmpRegions.size() << " objects - result " << m_regions.size() << std::endl;
 		}
 	}
@@ -324,7 +324,6 @@ void YoloDarknetDetector::Detect(const std::vector<cv::UMat>& frames, std::vecto
 	{
 		Detect(frames[0]);
 		regions[0] = m_regions;
-
 	}
 	else
 	{
@@ -351,11 +350,11 @@ void YoloDarknetDetector::Detect(const std::vector<cv::UMat>& frames, std::vecto
 					tmpRegions.emplace_back(cv::Rect(cvRound(wk * bbox.x), cvRound(hk * bbox.y), cvRound(wk * bbox.w), cvRound(hk * bbox.h)), T2T(bbox.obj_id), bbox.prob);
 			}
 
-			nms3<CRegion>(tmpRegions, regions[i], 0.4f,
+			nms3<CRegion>(tmpRegions, regions[i], static_cast<track_t>(0.4),
 				[](const CRegion& reg) { return reg.m_brect; },
 				[](const CRegion& reg) { return reg.m_confidence; },
 				[](const CRegion& reg) { return reg.m_type; },
-				0, 0.f);
+				0, static_cast<track_t>(0));
 		}
 
 		m_regions.assign(std::begin(regions.back()), std::end(regions.back()));

src/Detector/YoloTensorRTDetector.cpp

@@ -201,8 +201,8 @@ void YoloTensorRTDetector::Detect(const cv::UMat& colorFrame)
         std::vector<cv::Rect> crops = GetCrops(m_maxCropRatio, m_detector->GetInputSize(), colorMat.size());
         //std::cout << "Image on " << crops.size() << " crops with size " << crops.front().size() << ", input size " << m_detector->GetInputSize() << ", batch " << m_batchSize << ", frame " << colorMat.size() << std::endl;
         regions_t tmpRegions;
-		std::vector<cv::Mat> batch;
-		batch.reserve(m_batchSize);
+        std::vector<cv::Mat> batch;
+        batch.reserve(m_batchSize);
         for (size_t i = 0; i < crops.size(); i += m_batchSize)
         {
             size_t batchSize = std::min(static_cast<size_t>(m_batchSize), crops.size() - i);
@@ -212,7 +212,7 @@ void YoloTensorRTDetector::Detect(const cv::UMat& colorFrame)
 				batch.emplace_back(colorMat, crops[i + j]);
 			}
 			std::vector<tensor_rt::BatchResult> detects;
-            m_detector->Detect(batch, detects);
+			m_detector->Detect(batch, detects);
 
 			for (size_t j = 0; j < batchSize; ++j)
 			{
@@ -229,11 +229,11 @@ void YoloTensorRTDetector::Detect(const cv::UMat& colorFrame)
 
 		if (crops.size() > 1)
 		{
-			nms3<CRegion>(tmpRegions, m_regions, 0.4f,
+			nms3<CRegion>(tmpRegions, m_regions, static_cast<track_t>(0.4),
 				[](const CRegion& reg) { return reg.m_brect; },
 				[](const CRegion& reg) { return reg.m_confidence; },
 				[](const CRegion& reg) { return reg.m_type; },
-				0, 0.f);
+				0, static_cast<track_t>(0));
 			//std::cout << "nms for " << tmpRegions.size() << " objects - result " << m_regions.size() << std::endl;
 		}
     }

src/Detector/tensorrt_yolo/decodeTensorCUDA.cu

@@ -47,7 +47,7 @@ __global__ void decodeTensorKernel(
 float* decodeTensorCUDA(const int imageIdx, const TensorInfo& tensor)
 {
     // host
-    int boxes_bytes = 6*sizeof(float)*tensor.grid_h*tensor.grid_w*tensor.numBBoxes; // x y w h maxProb maxIndex 6个元素
+    int boxes_bytes = 6*sizeof(float)*tensor.grid_h*tensor.grid_w*tensor.numBBoxes; // x y w h maxProb maxIndex 6
 	const float* detections = &tensor.hostBuffer[imageIdx * tensor.volume];
     float* boxes = (float*) malloc(boxes_bytes);
 

src/Tracker/Kalman.cpp

@@ -17,12 +17,10 @@ namespace kalman = cv::detail::tracking;
 /// \param deltaTime
 /// \param accelNoiseMag
 ///
-TKalmanFilter::TKalmanFilter(
-        tracking::KalmanType type,
-	    bool useAcceleration,
-        track_t deltaTime, // time increment (lower values makes target more "massive")
-        track_t accelNoiseMag
-        )
+TKalmanFilter::TKalmanFilter(tracking::KalmanType type,
+                             bool useAcceleration,
+                             track_t deltaTime, // time increment (lower values makes target more "massive")
+                             track_t accelNoiseMag)
     :
       m_accelNoiseMag(accelNoiseMag),
       m_deltaTime(deltaTime),
@@ -374,7 +372,7 @@ void TKalmanFilter::CreateLinearAcceleration(cv::Rect_<track_t> rect0, Point_t r
 
     m_initialRects.reserve(MIN_INIT_VALS);
 
-	m_initialized = true;
+    m_initialized = true;
 }
 
 ///

src/common/defines.h

@@ -381,7 +381,7 @@ enum FilterGoal
 {
     FilterCenter,
     FilterRect,
-	FilterRRect,
+    FilterRRect,
     FiltersCount
 };