add native 3D object detection using ZED SDK optimized models

Add engines::ZedNative class wrapping ZED SDK's TensorRT-optimized
object detection. On every frame grab, detected objects (label,
confidence, 3D position, 2D/3D bounding boxes) are serialized as a
flat float32 array and sent as a new "objects" dora output with
frame_id correlation to image_bgr.

Extract shared env-var helpers into src/util/env.h and expose
sl::Camera& via Camera::handle() for engine access.

Closes #3

src/engines/zed_native.h

@@ -0,0 +1,41 @@
+#pragma once
+
+#include <sl/Camera.hpp>
+#include <string>
+#include <vector>
+
+namespace engines {
+
+struct DetectionConfig {
+    std::string model = "MULTI_CLASS_BOX_MEDIUM";
+    float confidence_threshold = 40.0f;
+    bool tracking = false;
+
+    static DetectionConfig from_env();
+};
+
+struct DetectedObject {
+    int label_id;
+    float confidence;
+    float position[3];        // center x, y, z (mm)
+    float bounding_box_2d[8]; // 4 corners × 2 coords (pixels)
+    float bounding_box_3d[24]; // 8 corners × 3 coords (mm)
+};
+
+class ZedNative {
+public:
+    explicit ZedNative(const DetectionConfig& config);
+
+    void enable(sl::Camera& camera);
+    void detect(sl::Camera& camera);
+    const std::vector<DetectedObject>& objects() const;
+
+private:
+    DetectionConfig config_;
+    sl::ObjectDetectionParameters det_params_;
+    sl::ObjectDetectionRuntimeParameters det_rt_params_;
+    sl::Objects raw_objects_;
+    std::vector<DetectedObject> detected_;
+};
+
+} // namespace engines