analyze_resolved()

Kinesis::Vision::VisionAnalysis MayaFlux::Yantra::VisionAnalyzer::analyze_resolved ( const std::shared_ptr< Core::VKImage > &  image )

private

Shared body of both analyze_vision() overloads, once a GPU image is in hand: resolve the configured query, run every resolved VisionSequence, apply context filtering, run the GPU reductions.

w/h are the image's own dimensions.

Definition at line 499 of file VisionAnalyzer.cpp.

{
    const auto resolved = Kinesis::Vision::resolve(m_query);
 
    std::vector<Kinesis::Vision::VisionResult> results;
    results.reserve(resolved.size());
    m_label_lane = 0;
 
    for (size_t lane = 0; lane < resolved.size(); ++lane) {
        if (!run_lane(lane, resolved, results, image))
            results.emplace_back();
    }
 
    auto analysis = assemble(resolved, results, image->get_width(), image->get_height());
    m_results = std::make_shared<const std::vector<Kinesis::Vision::VisionResult>>(std::move(results));
    return analysis;
}

References assemble(), image, m_label_lane, m_query, m_results, MayaFlux::Kinesis::Vision::resolve(), and run_lane().

Referenced by analyze_vision(), and analyze_vision().

Here is the call graph for this function:

Here is the caller graph for this function: