---
title: "Getting started"
slug: "getting-started-1"
updated: 2026-09-02T05:49:31Z
published: 2026-09-02T05:49:31Z
canonical: "docs-linescan.alliedvision.com/getting-started-1"
---

> ## Documentation Index
> Fetch the complete documentation index at: https://docs-linescan.alliedvision.com/llms.txt
> Use this file to discover all available pages before exploring further.

# Getting started

This chapter will describe the basic functions/sequences needed to handle the basic functionality of the camera.

Ready to use-Examples are also shipped with the SDK in order to demonstrate the usage of the SDK regarding getting/setting features and acquiring images.

## Initialization of the SDK

Before accessing any other functions of the SDK, an initialization needs to be done. Please refer to *Init/Deinit* -functions for the detailed description of the function **csiInit**.

After finishing the work with the SDK make sure to call the **csiClose** function. This makes sure that all memory is freed again, and all connections/interfaces are properly closed again.

## Connecting to a camera

The use of the Allied Vision Gen<I>CAM-SDK enables the user to use different transport layers and interfaces for the available devices.

Depending on the requirements for your application these transport layers can be selected during the device discovery process.

It is possible to use the standard search paths for the already installed transport layers.

These paths are set in the environmental variable *“GENICAM_GENTL64_PATH”* or for 32Bit-applications: *“GENICAM_GENTL32_PATH”* .

This is the default behavior. To reduce the time needed for the discovery process a specific path can be given. The search can also be limited to this single path when the *overrideSearchPath* is set.

To establish a connection, you will need to call 2 functions: **csiDiscoverDevices** and **csiOpenDevice**.

## Getting and setting features

To configure the camera, so called features can be set and read by using the feature names provided by the device-xml-file. All features are of a specific type. The following different types exist:

Boolean

Integer

Floating point

String

Command

Register

Enumeration

For each type, a “Get”- and “Set”-function does exist in the API. For example, use **csiGetFeatureFloat** to get a float parameter. To retrieve additional information, the function **csiGetFeatureParameter** exists. This function will fill a **csiFeatureParameter** structure which provides information about the display name, minimum and maximum values, etc. This function is especially useful if you do not know the valid thresholds of a parameter.

Please be careful when treating string features. You must not exceed the maximum length! This can also be retrieved with the function **csiGetFeatureParameter**. The parameter *maximumStringLength* of the **csiGetFeatureParameter** structure will indicate the maximum string length to set in the feature.

If the complete list of the device features needs to be retrieved, it is recommended to use the function *csiIterateFeatureTree* . An example is shipped with the SDK to demonstrate the usage of it.

To set the values, please use the type-specific set-functions. For example, use **csiSetFeatureInt** for an integer value.

## Acquiring images

To get images from the device, it must be opened first by calling the appropriate functions. The diagram below provides an overview of the functions which should be called during an acquisition process

![](https://cdn.document360.io/eca58697-9c2f-461e-9f0a-e6205b74b693/Images/Documentation/getting-started-image-pudfx8in.png)

**Image 1 overview_acquiring_process**

Depending on the type of the device, it is possible to retrieve multiple data streams in parallel from the same device. This capability can be checked by using the **csiGetNumberOfDataStreams** function.

In general, two different ways of acquiring the images can be used:

1. Using Events (Events must be registered by the **csiRegisterEvent** function prior to the usage of the event.)

2. Directly calling the **csiGetNextImage** function

Independent of these two ways, the acquisition from the device must be started first by calling **csiStartAcquisition**. If enough images have been processed, this needs to be stopped again by calling **csiStopAcquisition**. After a received image is processed, it must be released back into the receive buffer of the acquisition engine by calling **csiReleaseImage**. Failing to do so will cause an error as soon as all receive buffers have been filled by the incoming data.

To grab images continuously, the processing part needs to keep up with the speed of the camera. Otherwise, images might be lost.